Processing May 2013

Your search for expertise ends with us. We build better boilers.Satisfied customers discover the power of our practical knowledge – the ability to design and build boilers that operate efficiently, safely and cleanly in a variety of industrial applications, including refining, petro-chemical and power generation. The know-how of our engineers and technicians – combined with our expanded facilities and equipment, including a new membrane panel welding machine – results in economic value and competitive advantage for you. We’ve been designing and building boilers for people who know and care since 1996.

WWW.RENTECHBOILERS.COM

HydrocarbonProcessing.com | MAY 2013

®

MAINTENANCE AND RELIABILITY

Advances in monitoring/repair

methods and new equipment

designs increase unit availability

SAFETYInvestigation determines root

cause for flare header failure

PETROCHEMICAL DEVELOPMENTS

More efficient energy usage

increases profitability

for olefins production

Select 56 at www.HydrocarbonProcessing.com/RSSelect 56 at www.HydrocarbonProcessing.com/RS

Better performance, less downtime, better profitability – everyone approves of that. Call 1-866-335-3369 or visit sentron.ca to begin your trial.

Petro-Canada is a Suncor Energy businessTMTrademark of Suncor Energy Inc. Used under licence.

SENTRONTM LD 5000

Field Tested. Field Proven.

Select 76 at www.HydrocarbonProcessing.com/RS

Cover Image: Eastern Petrochemical Co. (SHARQ), a joint venture between SABIC and a Japanese consortium headed by Mitsubishi, is a world-scale polyethylene (PE)

complex. The facility produces 800,000 tpy of high-density PE (HDPE) and linear-low-density PE (LLDPE), in addition to the extrusion and automatic bagging and palletizing

lines. Linde Engineering Dresden GmbH fulfilled the construction contract for both PE processing operations (HDPE and LLDPE ), reactors and extrusion lines. Photo courtesy

of Linde Engineering Dresden GmbH, Dresden, Germany.

MAY 2013 | Volume 92 Number 5HydrocarbonProcessing.com

SPECIAL REPORT: MAINTENANCE AND RELIABILITY 39 Avoid hidden costs of suction-specific speed in pumping

J. Bailey and S. Bradshaw

43 Rerating rotating equipment optimizes olefins plant performance

D. Renard

49 Maximize steam unit performance with precise torque monitoring

T. Mayne, M. Ellul and D. Phillips

53 Investigate power limitations in a large steam turbine

N. Ghaisas

57 Seal safety may require going beyond typical standards

S. Shaw

61 Prevent methane hydrate formation in natural gas valves

A. Glaun and J. Shahda

69 Consider both actual and virtual spare parts inventory

H. P. Bloch

75 Improve design for pump suction nozzles

M. G. Choudhury, A. Kulkarni and D. Koranga

BONUS REPORT: PETROCHEMICAL DEVELOPMENTS 79 Consider novel CGC and front-end depropanizer system

for olefins production

J. Fu, C. Zhao and Q. Xu

85 Enhance operation and reliability of dividing-wall columns

J. Shin, J. Lee, S. Lee, B. Lee and M. Lee

TERMINALS AND STORAGE—SUPPLEMENT 92 Overflow systems are the last line of defense

M. Toghraei

94 Terminals and storage news

REFINING DEVELOPMENTS 99 Optimize vacuum ejector operations

T. Temur, M. Haktanir, F. Uzman, M. Karakaya and A. K. Avci

SAFETY/LOSS PREVENTION 105 Flare header failure: An investigation

J. Tharakan

HP ONLINE EXCLUSIVE Innovations

DEPARTMENTS

4 Industry Perspectives

8 Brief

11 Impact

17 Associations

110 Marketplace

113 Advertiser index

COLUMNS

21 Reliability

Manage your time constructively

25 Integration Strategies

Asset information management improves life cycle benefits for equipment

29 Boxscore Construction

Analysis

Jurong Island—Asia’s preemptive LNG trading hub?

35 Viewpoint

Consider integral-gear compressors in CO2 services

114 Engineering Case Histories

Case 72: Interaction between disciplines when troubleshooting

78

38

4�MAY 2013 | HydrocarbonProcessing.com

P. O. Box 2608Houston, Texas 77252-2608, USAPhone: +1 (713) 529-4301Fax: +1 (713) 520-4433Editorial@HydrocarbonProcessing.comwww.HydrocarbonProcessing.com

President/CEO John Royall

Vice President Ron Higgins

Vice President, Production Sheryl Stone

Business Finance Manager Pamela Harvey

Part of Euromoney Institutional Investor PLC. Other energy group titles include: World Oil and Petroleum Economist

Publication Agreement Number 40034765 Printed in USA

Industry PerspectivesPUBLISHER Bret Ronk [email protected]

EDITORIAL Editor Stephany RomanowReliability/Equipment Editor Heinz P. BlochManaging Editor Adrienne BlumeTechnical Editor Billy ThinnesOnline Editor Ben DuBoseAssociate Editor Helen MecheDirector, Data Division Lee NicholsContributing Editor Loraine A. HuchlerContributing Editor William M. GobleContributing Editor ARC Advisory Group

MAGAZINE PRODUCTIONVice President, Production Sheryl StoneManager, Editorial Production Angela BatheArtist/Illustrator David WeeksGraphic Designer Amanda McLendon-BassManager, Advertising Production Cheryl Willis

ADVERTISING SALESSee Sales Offices, page 113.

CIRCULATIONDirector, Circulation Suzanne McGehee +1 (713) 520-4440 [email protected] price (includes both print and digital versions): Print—One year $239, two years $419, three years $539. Digital format—One year $239. Airmail rate outside North America $175 additional a year. Single copies $35, prepaid.

Because Hydrocarbon Processing is edited specifically to be of greatest value to people working in this specialized business, subscriptions are restricted to those engaged in the hydrocarbon processing industry, or service and supply company personnel connected thereto.

Hydrocarbon Processing is indexed by Applied Science & Technology Index, by Chemical Abstracts and by Engineering Index Inc. Microfilm copies avail-able through University Microfilms, International, Ann Arbor, Mich. The full text of Hydrocarbon Processing is also available in electronic versions of the Business Periodicals Index.

ARTICLE REPRINTSIf you would like to have a recent article reprinted for an upcoming confer-ence or for use as a marketing tool, contact Foster Printing Company for a price quote. Articles are reprinted on quality stock with advertisements removed; options are available for covers and turnaround times. Our minimum order is a quantity of 100.

For more information about article reprints, call Rhonda Brown with Foster Printing Company at +1 (866) 879-9144 ext. 194 or e-mail [email protected].

Hydrocarbon Processing (ISSN 0018-8190) is published monthly by Gulf Publishing Company, 2 Greenway Plaza, Suite 1020, Houston, Texas 77046. Periodicals postage paid at Houston, Texas, and at additional mailing office. POSTMASTER: Send address changes to Hydrocarbon Processing, P.O. Box 2608, Houston, Texas 77252.

Copyright © 2013 by Gulf Publishing Company. All rights reserved.

Permission is granted by the copyright owner to libraries and others registered with the Copyright Clearance Center (CCC) to photocopy any articles herein for the base fee of $3 per copy per page. Payment should be sent directly to the CCC, 21 Congress St., Salem, Mass. 01970. Copying for other than personal or inter-nal reference use without express permission is prohibited. Requests for special permission or bulk orders should be addressed to the Editor. ISSN 0018-8190/01.

US-based ultra-low-sulfur gasoline rule released

In late March, the US Environmental Protection Agency (EPA) released the long-awaited ruling for ultra-low-sulfur gasoline, also known as the Tier 3 rule. This proposed rule lowers the sulfur content of gasoline by 60% from the present level of 30 ppm to 10 ppm by 2017. The goal of Tier 3 is to prevent 2,400 premature deaths per year and 23,000 cases of respiratory ailments in children, all totaling $8 billion and $23 billion in health-related benefits. It supports efforts by states to reduce harmful levels of smog and soot sourced from vehicle emissions. The new gasoline sulfur rule has support from state governors, public health groups, environmentalists and the auto industry.

The US Tier 3 rule follows a similar 10-ppm specification set by the European Union’s Euro V enacted in 2009. EPA con-tends that the refining industry can achieve these quality speci-fications with available technologies and at reasonable costs.

Counterpoint. The American Petroleum Institute (API) and the American Fuel and Petrochemical Manufacturers (AFPM) offer different opinions. EPA’s proposed Tier 3 fuel regulations could raise refiners’ costs, provide little or no environmental ben-efit, and actually increase carbon emissions, according to API.

“There is a tsunami of federal regulations coming out of the EPA that could put upward pressure on gasoline prices. EPA’s proposed fuel regulations are the latest example. Consumers care about the price of fuel, and our government should not be adding unnecessary regulations that raise manufacturing costs, especially when there are no proven environmental benefits. We should not pile on new regulations when existing regulations are working,” said to API Downstream Director, Bob Greco.

According to API, EPA’s Tier 3 proposal would increase the cost of gasoline production by up to 9¢/gal based on an analysis by the energy consulting firm, Baker & O’Brien. If EPA adds a vapor-pressure reduction requirement in a separate regula-tion, it would push the cost increase up to 25¢/gal, according to Baker & O’Brien.

“Tier 3 rulemaking that targets trace amounts of sulfur in gasoline is not worth the direct threat to our domestic fuel sup-ply, consumer cost at the pump and American jobs,” accord-ing to AFPM President Charles T. Drevna. US refiners have already spent billions of dollars to achieve a 90% reduction in sulfur levels, but Tier 3 will require another $10 billion in new infrastructure and another $2.4 billion/yr in operating costs.

Balance. Tier 3 specifications will increase greenhouse gas (GHG) emissions to produce the 10-ppm sulfur gasoline. The goal of Tier 3 is to improve air quality. Yet, the path to Tier 3 will generate more GHGs to process crude oil into ultra-low-sulfur transportation fuels.

An expanded version of Industry Perspectives can be found online at HydrocarbonProcessing.com.

IT’S S

AFE

innovate/customize/educate


NEW DELHI, INDIA | 9–11 JULY

Program Print Sponsor:

Speaker Gift Sponsor:

Conference Lanyard Sponsor:

EVENT

Conference Host:

Technical Program Sponsor:

Refi ning Track Sponsor:

Conference Delegate Bag Sponsor:

Media Sponsors:

THE AUTHORIT Y ON ENERGY

Petrochemical Track Sponsor:

IRPC 2013: At The Forefront of New Technology and Operations Advancements at Refi neries and Petrochemical PlantsJoin your peers and leaders from the HPI industry in a high-level discussion and look at the latest advances in technology and operations at refi neries and petrochemical plants.

This year, advancements in technology and operations will be explored, including the unique opportunities found in greater refi nery and petrochemical integration. During the two days of technical sessions, well-respected industry speakers will give case studies, examples and insight into technology and trends that are revolutionizing the HPI.

Agenda Highlights Include:

View the full list of speakers and the complete conference agenda online at HPIRPC.com

TRACK 1 HIGHLIGHTS:REFINING

TRACK 2 HIGHLIGHTS:PETROCHEMICALS

Profi t Improvement Program in KNPC Refi neries - Abbas Shamash, Kuwait National Petroleum Company

Optimization of Olefi n Plants - Vera Varaprasad, Indian Oil Corporation Limited

Challenges in Design and Engineering of High Pressure Hydrotreaters and Avenues for Energy Optimization - Mr. K. Sudhaker, L&T Chiyoda Ltd.

Case Study: Design, Development, and Deployment of Energy Management System (ISO-50001:2011) at an Integrated Petrochemicals Complex - Mayur Talati, Reliance Industries Ltd.

Petrobras ULSD Revamps - Silvio Jose Vieira Machado, Petrobras Challenges and Emission Reduction in Integrating Refi neries and Petrochemical Plants, Romel Bhullar, Fluor Corporation; Anil Rajguru and Les Antaff y, Fluor Enterprises

KEYNOTE SPEAKER:

R.K. Ghosh, Director (Refi neries), Indian Oil Corporation Limited (Refi neries Division)

KEYNOTE SPEAKER:

Dr. Ajit Sapre, Group President, Research and Technology, Reliance Technology Group

Vco

HPIRPC.com

8:30-9:15 a.m. CONTINENTAL BREAKFAST9:15-9:30 a.m. OPENING REMARKS: Stephany Romanow, Editor, Hydrocarbon Processing

KEYNOTE SPEAKERS9:30-10:45 a.m. Dr. Ajit Sapre, Group President, Research and Technology, Reliance Technology Group; ExxonMobil Singapore (invited)

10:45-11 a.m. COFFEE BREAKTRACK 1 - REFINING

TRACK 2 - PETROCHEMICALS

11 a.m.-1 p.m. SESSION 7Session Chair: K.Venkataramanan, CEO and Managing Director, Larsen ToubroSESSION 8 Session Chair: John Baric, Licensing Technology Manager, Shell Global

Solutions International B.V.

Petrobras ULSD Revamps - Silvio Jose Vieira Machado, PetrobrasPhyrophoric Hazard of Catalyst Handling in Refining and Petrochemical

Industry - Renato Benintendi, Foster Wheeler

Advanced Techniques for Enhancing Hydrogen Availability - Sanjiv Ratan,

Technip Stone & Webster Process Technology Economic Catalysis Solution for Acetylene and MAPD Selective

Hydrogenation - Shankhaneel Borah, Sud Chemie India Limited,

a Clariant Group Company

Improve Profitability by Flexibility Hydrocracking Technologies - Li Hui Ng -

CRI/Criterian Marketing Asia Pacific Pte. LTD Extending the Performance of Maximum Propylene Catalyst and Additives

- Vipan Goel, W.R. Grace

Recovery of Valuables from Refinery Off Gases to Increase Profit

Margins - Siddartha Murkerjee, Air Liquide Global E&C Solutions

Indiate Private Limited Comparative Study of Conventional Petrochemicals (Ethylene Glycol) With the

Bio Based (Bio-Ethylene Glycol) Production with the Application of Life Cycle

Methodology and Foot Printing Tools - Sharma Rajeev Kumar, India Glycols Limited

1-2 p.m.LUNCH

2:00 - 2:30 COFFEE & DESSERTS - EXHIBIT HALLTRACK 1 - REFINING

TRACK 2 - PETROCHEMICALS

2:30-4:30 p.m. SESSION 9 Session Chair: B.K. Namdeo, Executive Director-International Trade

& Supplies, Hindustan Petroleum Corporation LimitedSESSION 10 Session Chair: Chakrapany Manoharan, Director-Refinery, Essar Oil, Ltd.

Energy Consumption Update in Amuay Refinery of Paraguaná Refining

Center (CRP) - Daniel Reyes, PDVSAEDC Pyrolysis Furnace Radiant Section Tube Failure-Case Study

- Mr. K. Ramesh, Reliance Industries

Profit Improvement Program in KNPC Refineries - Abbas Shamash, Kuwait

National Petroleum CompanyTemperature Dependent Catalysts: Optimizing Performance and ROI

with Advanced Temperature Measurement Systems - Walter Tijmes,

Daily Thermetrics Corporation

A Study of Mechanical Failure of Spent Acid Regeneration Combustion

Chamber - S. Saha, ENGG DIV. RPTL (JEC). Reliance Refinery, Jamnagar, IndiaControlling Corrosion in Process Refrigeration Systems and Gas Compression

Packages - Amey Majgaonkar, Kirloskar Pneumatic Co. Ltd.

Refinery Operation: The Same Tool Optimizes Both Mid Term Planning

& Scheduling – Aurelio Ferrucci, PROMETHEUS S.r.l. Best Practices for Mitigation of Corrosion in Hydrocarbon Processing Industry

- K. Sudhakar, L&T Chiyoda Limited

4:30-5 p.m. AFTERNOON BREAKTRACK 1 - REFINING

TRACK 2 - PETROCHEMICAL

5-6:30 p.m. SESSION 11 Session Chair: Stephany Romanow, Editor, Hydrocarbon Processing SESSION 12 Session Chair: A.K. Purwaha, Chairman and Managing Director,

Engineers India, Ltd.

Analysis of FCC Reactor Cyclone Flow - Narasimhamurthy Bontu,

Reliance Industries, Ltd.

Virtual Reality as Effective Tool for Training Field Operators and Making

Decisions: Experiment Results - Alessandro Altamura, Technip Italy S.p.A.

Sour Water Stripper Units with High Cyanides Contents - Vikas Kapoor,

Fluor Daniel India Pvt. Ltd.E-Learning and Universal Simulation for Competency Development -

Mr. Santosh Joshi , GSE EnVision

Coke Drum Unheading - Elango Narendran, DeltaValveRegaining Operating Excellence Through Enhanced Training

- Michael A. Taube, S&D Consulting LLC

6:30-8 p.m. CLOSING RECEPTION

IRPC 2013 Agenda Day 2 | Thursday, 11 JulyFASTOPENING REMARKS: Stephany RomanoKEYNOTE SPEAKERS9:30-10:45 a.m. Dr. Ajit Sapre, Group President, Research and TechnologEE BREAKK 1 - REFINING

N 7 KK.Vennkatararamanan,n, CEO and Managing DirectoRevamps - Silvio Jose Vieira Machado, Petroues for Enhancing Hydrogen Availability - ebsbsterr Prococess Technhnology

by Flexibility Hydrocracking Technologiesg AAsia PPacifiific Pte. LTTDrom Refinery Off Gases to Increase Profit kerjee, Air LLiquide GGlobal E&C Solutions

RTS - EXHIBIT HALLG

utivive DDirectotor-Internrnational Trade orpporaration n Limiteday Refinery of Paraguaná Refining Refineries - Abbas Shamash, Kuwai

Regeneration Combustion eliaance e Refinnery, Jammnagar, Indias Both Mid Term Planning S.rS.r.l.

T

bonon Prorocessssing SSes

EngBontu, Virtu

DecisVikas Kapoor, E-LearMr. San

Regainin- Michae

HPIRPC.com

8:30-9:15 a.m. CONTINENTAL BREAKFAST

9:15-9:30 a.m. OPENING REMARKS: John Royall, President and CEO, Gulf Publishing Company

9:30-10:45 a.m. KEYNOTE SPEAKERS

R.K. Ghosh, Director (Refineries), Indian Oil Corporation Limited (Refineries Division)

Future Challenges and Opportunities for Refining in Asia - Suresh Sivanandam, Wood Mackenzie

10:45-11 a.m. COFFEE BREAK

TRACK 1 - REFININGTRACK 2 - PETROCHEMICALS

11 a.m.-1 p.m. SESSION 1

Session Chair: Syamal Poddar, President, Poddar and Associates

SESSION 2

Session Chair: Eric Benazzi, Marketing Director, Axens

Diesel from Waste Plastics: Use in Vehicles - SK Singal, CSIR-Indian Institute

of Petroleum

Catalytic Olefins Technology Enhances Olefins Producers’ Flexibility

and Economics - Sourabh Mukherjee, KBR

Liquid Fuel from Coal – New Horizons - Atul Choudhari, Aker Powergas

PvtLtd, Mumbai

Optimization of Olefin Plants - Vera Varaprasad,

Indian Oil Corporation Limited

Hydroprocessed Diesel Produced as Byproduct of Bio-Jet Fuel Process:

A Superior Fuel in IC Engines - M.O. Garg, CSIR-Indian Institute

of Petroleum

Maximizing pX Production Through Optimizing the Phenyl-methyl Group

- Joseph C. Gentry, GTC Technologies US LLC

Minimizing Impact on Carbon Footprint while Processing Opportunity

Crudes - Tanmay Taraphdar, Technip KT India Limited

Polymer Process Developments and Second Generation Metallocene

Catalysis - Dr. Howard Paul, P.E., SK E&C USA, Inc.

1-2 p.m. LUNCH

2-2:30 p.m. COFFEE & DESSERTS - EXHIBIT HALL


2:30-4:30 p.m. SESSION 3

Session Chair: P.P. Upadhya, Managing Director, MRPL

SESSION 4

Session Chair: Giacomo Fossataro, General Manager, Walter Tosto S.p.A

Axens Solutions for Middle Distillate Hydroprocessing: A Focus on Revamping

and New Catalyst Technology - Stefania Archambeau, Axens

Challenges and Emission Reduction in Integrating Refineries and

Petrochemical Plants, Romel Bhullar, Fluor Corporation; Anil Rajguru and

Les Antaffy, Fluor Enterprises

Implementation of an Integrated Refinery Complex Production

Reconciliation: Benefits and Challenges - Srinivas Badithela, HMEL

Economics-Refinery/Petrochemical Integration - Sanjiv Singh, Panipat

Refinery and Petrochemical Complex, Indian Oil Corporation Limited

Effect of Reliability on Return of Invested Capital ROIC

- Logan Anjaneyulu, Valero

Dynamic Simulation: An Efficient Tool for Verifying Plant Integrity and

Control System Design - Sheo Raj Singh, Engineers India Limited

Challenges in Design and Engineering of High Pressure Hydro Treaters and

Avenues for Energy Optimization - Mr. K. Sudhaker, L&T Chiyoda Ltd.

Optimum Isolation Valve Sealing on Black Powder-Generated Gas Processing

- Omar Al Amri, Saudi Aramco

4:30-5 p.m. AFTERNOON BREAK


5-6:30 p.m. SESSION 5

Session Chair: A.S. Basu, Managing Director, Chennai Petroleum Corporation Ltd.SESSION 6

Session Chair: Carlos Cabrera, Executive Chairman, Ivanhoe Energy

New Solutions from Eco-friendly Gasoline Production - Adarsh Tripathi,

RRT Global

Maximizing Operational Efficiency and Safety by the Use of a Digital Plant -

Sloane Whiteley, AVEVA

Overcoming Pressure Drop Limitations in Hydroprocessing Reactors -

Mahendranadh Desu, Hindustan Petroleum Corporation Ltd.

Case Study: Design, Development and Deployment of Energy Management

System (ISO-50001 : 2011) at an Integrated Petrochemicals Complex - Mayur

Talati, Reliance Industries Ltd.

Optimum Refinery for Changing Feed and Product Demands - Samir

Saxena, KBR

Computational Fluid Dynamics as an Emerging Tool to Improve the Reliability

of the Plant Operations - S. Sathish Kumaran, Technip KT India Limited

6:30-8 p.m. CLOSING RECEPTION - Exhibit Hall

IRPC 2013 Agenda Day 1 | Wednesday, 10 July


EARLY BIRD SPECIAL

Register by 31 Mayto Save!

Supported by:

Register at HPIRPC.com by 31 Mayto SAVE UP TO 20%

Take part in this innovative, industry-leading forum. Register today and benefi t from:• More than 40 technical sessions given by speakers representing companies like Indian Oil Corporation Limited, Valero, Saudi Aramco, Hindustan Petroleum Company, Reliance Refi nery, PDVSA, Technip and more

• Join international HPI professionals from around the world, representing operators, refi neries and petrochemical plants like BP plc, Chevron Lummus Global LLP, ConocoPhillips Ltd, eni, ExxonMobil Research & Engineering Company, Linde Gas, Lukoil and Total

• Ample networking opportunities between sessions allow you to connect with old and new business contacts

• Explore and learn more about the latest developments in operations and technology

• Get a local and global perspective on the refi ning and petrochemicals industries

Conference registration includes:• Pre-conference tour of Indian Oil’s Panipat Refi nery & Petrochemicals Complex (9 July 2013)*

• Two-day conference program (10-11 July 2013) including keynote addresses, general presentations and panel discussions

• Breakfasts, lunches and refreshment breaks

• Access to exhibition fl oor throughout conference activity *Refi nery tour registration is offered on a fi rst-come, fi rst-served basis. Space is limited.

2013 Conference Fees:

Early Bird Fee (by 31 May) Regular Fee

Individual $945 $1,045

Team of Two $1,700 $1,875

Group of Five $4,250 $4,700

For more information about IRPC 2013, please contact Melissa Smith, Events Director, Gulf Publishing Company, at +1 (713) 520-4475 or [email protected].

* IRPC 2013 is currently the only refi ning and petrochemical event that will be held in India during 2013. Attendees will have the opportunity to take part in ane exclusive tour of Indian Oil’s Panipat Refi nery & Petrochemicals Complex, located 23 kilometers from Panipat and approximately 100 kilometers from New Delhi. Since its original construction, the Panipat refi nery has undergone numerous upgrades and unit additions, to arrive at its current capacity of 15 MMtpy.

| Brief

Centrica signs long-term North American LNG export contract with CheniereCentrica has an agreement with Cheniere Energy Partners to purchase 89

billion cubic feet of annual liquefied natural gas (LNG) volumes for export

from Cheniere’s Sabine Pass liquefaction plant in Louisiana. This amounts to

approximately 1.75 MM metric tpy, and is the equivalent of the annual gas

demand of around 1.8 MM UK homes. The contract is for an initial 20-year

period, with the option for a 10-year extension, and the target date for first

commercial delivery is September 2018. Under the terms of the agreement,

Centrica will purchase LNG on a free on board (FOB) basis, giving it destination

rights for the cargoes, for a purchase price indexed to the Henry Hub natural

gas price plus a fixed component. Centrica will export gas from the fifth LNG

train at Sabine Pass, on which preliminary engineering work has already begun.

Upon receiving news of the agreement, UK Prime Minister David Cameron said,

“I warmly welcome this commercial agreement between Centrica and Cheniere.

Future gas supplies from the US will help diversify our energy mix and provide

British consumers with a new long-term, secure and affordable source of fuel.”

The natural gas Killingholme Power Station, located in North Lincolnshire, UK,

will start using LNG from the US Gulf Coast sometime in 2019.

Hydrocarbon Processing | MAY 2013�9

BILLY THINNES, TECHNICAL EDITOR / [email protected]

Brief

The Center for Chemical Process Safety (CCPS) of the American Institute of Chemical Engineers (AIChE) has signed an agreement with India’s Oil Industry Safety Directorate (OISD). The agreement calls for the groups to cooperate in advancing the practice of process safety in the petroleum sector in India and around the world. The signing ceremony took place in March during a CCPS workshop on recognizing hazardous incident warning signs in Goa, India. The agreement broadens the sharing of lessons learned from process safety incidents, and includes information on the latest developments in the practice of process safety.

At the request of the National Institute of Standards and Technology (NIST), representatives from the Automation Federation participated in the first NIST meeting for developing a US cyber security program, as requested by US President Barack Obama. The meeting was held at the US Department of Commerce in Washington, DC. It is an important step in establishing the cyber security framework within President Obama’s executive order announced in his State of the Union address to confront the growing threat of cyberattacks on the US’ critical infrastructure. The framework will include “standards, methodologies, procedures and processes that align policy, business and technological approaches to address cyber risks,” and “help owners and operators of critical infrastructure identify, assess and manage cyber risk.” The next NIST cyber security framework meeting is scheduled for May 29 at Carnegie Mellon University in Pittsburgh, Pennsylvania.

US-based oil and gas industry professionals are the sixth best rewarded in the world, according to a global salary guide produced annually by Hays Oil and Gas and Oil and Gas Job Search. The guide reveals that industry professionals working in the US earn an average of $121,400 per year. This is slightly behind Canada at $123,000 per year. Australia is now the oil and gas pay leader, with an average industry salary of $163,600 per year. Expat industry workers are enjoying the highest levels of pay in Australia, the Philippines and Trinidad. Employer confidence in increasing staffing levels during 2013 remains high, with 75% of respondents predicting continued hiring increases. The guide bases its figures on more than 25,000 industry professionals across the globe.

ExxonMobil Pipeline Co. reported a leak in its Pegasus pipeline on March 29, resulting in the release of 3,500 to 5,000 barrels of crude outside of Mayflower, Arkansas. The pipeline—which runs from Patoka, Illinois to Nederland, Texas—was carrying Canadian Wabasca heavy crude at the time of the leak. The company says progress continues in the cleanup areas. Approximately 640 people are responding to

the incident in addition to federal, state and local responders. Cleanup has continued 24 hours a day since the spill was first detected. ExxonMobil also indicated that nearby Lake Conway remains oil free and stated that a comprehensive containment system using a boom has been deployed as a precaution. Regarding the pipeline, an excavation and removal plan for the affected portion of the pipeline is being developed for review by the US Department of Transportation. ExxonMobil has also received a corrective action order from the US Pipeline and Hazardous Materials Safety Administration.

PetroChina recently announced 2012 earnings of RMB 115.3 billion, which was a drop of over RMB 17 billion compared with 2011 earnings. The company’s earnings decline could be attributed to high natural gas import prices, as well as government caps on domestic fuel prices that eroded refining margins. PetroChina saw strong upstream growth in 2012, but the downstream side of its operations operated at an RMB 43.5 billion loss. Still, this was a slight improvement over 2011 downstream losses, which were RMB 61.9 billion. The continued downstream sector losses were due to PetroChina’s inability to shift the burden of rising oil costs to its consumers, as Chinese state policy mandates a cap on gasoline and diesel prices. In 2012, PetroChina’s refinery division produced 91 MM tons of gasoline, diesel and kerosine, an increase from the 87.2 MM tons in 2011. The company also produced 6 million tons of synthetic resin in 2012 (a rise of 7% year over year) and manufactured 3.7 MM tons of ethylene (up 6.4% from 2011).

The Organization of the Petroleum Exporting Countries (OPEC) sent a letter of condolence to acting Venezuela President Nicolás Maduro Moros following the death of the country’s President, Hugo Chávez. The president of the OPEC Conference, Hani Abdulaziz Hussain, who is also Kuwait’s Minister of Oil, referred to President Chávez’s “long and brave battle with cancer” and, on behalf of the conference, conveyed his “sincerest condolences” to President Maduro and, through him, to the people of Venezuela. He added: “There can be no doubt that President Chávez will be long remembered for his strong support to OPEC.”

Shell plans to sell its refinery in Geelong, Australia. The proposed sale of the 120,000-bpd refinery is in line with Shell’s global strategy to concentrate investment on large-scale sites, such as the company’s Pulau Bukom refinery in Singapore. The announcement further underpins Shell’s Australia strategy to grow its retail and bulk fuels business, along with terminals and pipelines. The company aims to conclude the sales process by the end of 2014. “This announcement will be difficult for refinery employees, but Shell will support them,” a Shell executive said.

SUPERIOR SPRAY. SERIOUS RESULTS. Whether you need to cool gas, dissolve salts in an overhead line or inject chemicals to prevent corrosion, we can help optimize injector performance. Here's how:

• Assistance with nozzle selection, spray direction and injector placement. There are dozens of factors to consider before choosing a spray nozzle, determining whether to spray co- or counter-current and identifying the proper placement of an injector in a vessel. We can help you evaluate your process conditions and then design an injector to provide optimal performance

• Design validation using Computational Fluid Dynamics (CFD) and Fluid Structure Interaction (FSI). We use powerful modeling tools to simulate your environment, confirm the injector will provide the expected spray performance and withstand process conditions such as thermal stresses, heat transfer, vortex shedding and more

• Proven track record. Companies like Technip, Mustang Engineering, Bechtel, Shell and many others rely on us to manufacture B31.1 and B31.3 code-compliant injectors and conduct radiographic, hydrostatic, ferrite tests and more

Learn More. Call 1.800.95.SPRAY or visit spray.com/injectors

This injector sprays liquid into gas and is just one of many used for gas cooling, water wash, desuperheating, steam quench, slurry backflush and more.

WIDE RANGE OF HYDRAULIC & GAS ATOMIZING NOZZLES INCLUDING CLOG-RESISTANT STYLES

DOZENS OF INJECTOR DESIGNS AVAILABLE

Unmatched Global Engineering, Manufacturing & Technical Support Nozzles | Control Systems | Headers & Injectors | Research & Testing

1.800.95.SPRAY

CFD MODEL ILLUSTRATES PERFORMANCE BASED ON INJECTOR PLACEMENT

ZENS OF INJECTORSIGNS AVAILABLEUDING

STANT STYLE



Impact

BILLY THINNES, TECHNICAL EDITOR / [email protected]

Global oil flows shift from West to East

By 2020, crude oil trade flows west of Suez will drop by 4.2 MM bpd, while crude flows east of Suez will rise by 4.7 MM bpd (FIG. 1), according to a recently published report from ESAI Energy. The primary global implication of the growth in US shale liquids and Canadian oil sands is significant change in crude oil trade flows. These changes will funda-mentally alter the relationship between OPEC countries and the consuming countries of North America, Europe and Asia-Pacific. Relative to the US and Eu-rope, the countries of Asia-Pacific will become far more dependent on OPEC countries for oil supplies. This will force the countries of Asia, especially China, to take a more deliberate role in responding to Middle Eastern conflicts and instabili-ty in producing countries, which, in turn, will alter the relationship of these coun-tries to the US, the current guarantor of the sea lanes.

The report outlines some of the po-tentially most significant changes to crude flows, including the reduction of North American imports from Latin America, Africa and the Middle East by 3 MM bpd between 2012 and 2020, and a drop in European imports from the same regions of another 2 MM bpd. As this happens, crude oil producers will in-creasingly focus on Asia-Pacific custom-ers. FSU, African and Latin American exports to the region will rise by 1.7 MM bpd, theoretically leaving the remaining market for Middle Eastern supplies of 3 MM bpd or an annual increment of as much as 375,000 bpd.

“There are many implications of these initial trends in global oil trade,” said Sarah Emerson of ESAI Energy. “US and European interest in committing re-sources and personnel to the security of the Persian Gulf may face renewed politi-cal resistance. Competition for the Asia-Pacific market is bound to weaken crude prices, and China’s disproportionate de-

pendence on imported oil will hasten ef-forts to improve energy security, includ-ing the inevitable development of shale.”

UK consumers buying less fuel at the pump

In the last 10 years, average UK fuel prices have almost doubled, rising from 73.68 p/l to 136.26 p/l for unleaded gasoline, and 75.57 p/l to 142.39 p/l for diesel. The latest retail marketing survey, conducted by the Energy Institute (EI), shows there are more registered UK vehi-cles on the road than ever before, yet total fuel sales have dropped by 6% since 2002.

A cutback in fuel sales suggests im-provements in engine performance and fuel economy, combined with changes in driver behavior. This is supported by die-sel sales outperforming gasoline for the

second year running. The number of gas stations in the UK stood at 8,693 at the close of 2012. This is compared to 1967’s all-time high of 39,958. Key findings of the survey show:

• Gasoline sales fell marginally to 13.42 MM tons by the close of 2012, down from 13.86 MM at the end of 2011

• Diesel sales totaled 13.86 MM tons by year end, down slightly from 13.91 MM tons in 2011

• Total 2012 road fuel sales fell slightly to 35.35 MM tons

• By the close of 2012, unleaded gasoline prices had averaged 136.26 p/l (vs. 133.6 p/l in 2011), while diesel pric-es closed the year at an average price of 142.39 p/l (vs. 138.90 p/l in 2011)

• Registered UK vehicles once again broke records, rising from 34.67 MM in 2011 to reach 36.71 MM by the end of

0.1 to 0.1

0.1 to 0.1 4.2 to 5.0

0.3 to 0.4 1.1 to 1.7

12.2 to 15.2

1.9 to 1.12.7 to 1.41.0 to 0.1

0.5 to 0.5

2.2 to 1.0

2.7 to 1.8

1.1 to 1.6

2.2 to 2.8

0.5 to 0.4

2012 flows west of Suez: 16.1 MMbpd2020 flows west of Suez: 11.9 MMbpd

2012 flows east of Suez: 16.6 MMbpd2020 flows east of Suez: 21.3 MMbpd

Source: ESAI Energy, March 2013

NorthAmerica

LatinAmerica

Africa Middle East Asia-Pacific

East EuropeEurope

FSU

FIG. 1. Approximate changes in net crude, 2012 to 2020 (MMbpd).

TABLE 1. US corrosion inhibitor demand (MM dollars)

% Annual growth

Item 2007 2012 2017 2007–2012 2012–2017

Corrosion inhibitor demand 2,070 2,030 2,485 –0.4 4.1

Petroleum refi ning 500 540 640 1.6 3.5

Utilities 375 343 390 –1.8 2.6

Chemicals 361 310 385 –3 4.4

Oil and gas production 182 338 475 13.2 7

Other 652 499 595 –5.2 3.6

Impact

12

2012, with each gas station servicing an average of 3,993 vehicles.

The survey also took care to break down the differences in gas station sites by sector for 2012:

• Oil company sites decreased by 151 to 5,159

• Main retailer sites increased by 423 to 1,233

• Supermarket sites increased by sev-en to 1,317

• Smaller retailer sites increased by two to 62

• Other unbranded sites increased by 37 to 922.

The four largest oil company opera-tions by number of branded gas stations were (2011 figures in brackets):

• BP, 1,220 (1,178) up 42• Shell, 1,028 (845) up 183• Esso, 907 (890) up 17• Texaco, 787 (840) down 53.

US demand for corrosion inhibitors to reach $2.5 billion

US demand for corrosion inhibitors is forecast to rise 4.1% per year to $2.5 billion in 2017, with volume demand approaching 1.7 billion pounds (TABLE 1). Growth will be driven by higher oil and natural gas output, particularly from shale formations, as well as by increasing chemical production and an expanding economy. Additionally, robust increases in construction spending will support demand for corrosion inhibitors used in cement and concrete, industrial coatings and metal applications. Value growth will also be aided by the introduction of new hybrid products that have functions in addition to corrosion protection. These and other trends are presented in a new study from The Freedonia Group.

The oil and gas industry’s continued expansion of horizontal drilling and hy-drofracturing well stimulation in shale formations will drive increases in cor-rosion inhibitor demand going forward, especially organic inhibitors. Increas-ingly caustic water produced by exist-ing oil wells will support higher organic inhibitor usage rates, as will efforts to reuse and recycle water to avoid addi-tional freshwater use. The availability of relatively cheap natural gas will spur faster growth in chemical production, leading to advances in corrosion inhibi-tor demand in both water treatment and process additive applications.

The fastest growth in corrosion in-hibitor demand, albeit from a small base, will occur in concrete and cement additives due to a rebound in construc-tion spending. Nitrites will benefit due to their popularity for protecting metal rebar in reinforced concrete. Higher construction spending will also sup-port demand for corrosion inhibitors in industrial coatings, particularly as im-provements in state and local finances allow for greater spending on infrastruc-ture maintenance and modernization.

In a number of more mature mar-kets such as petroleum refining, metals and utilities, moderate growth will be supported by an expanding economy. Water treatment corrosion inhibitors ac-count for the greatest share of demand in these markets, though in most cases process and product corrosion inhibi-

VEGA offers a complete line of level measurement instruments using radiation-based technology for accuracy in the most extreme environments.

www.vega-americas.com

[email protected]

1-800-FOR-LEVEL

VEGA detectors supply the following benefits:

Non-contact measurement is unaffected by corrosive process properties and internal obstructions

Patented flexible detector follows vessel contours and avoids obstructions

External mounting reduces both short and long term maintenance costs

Complete Level Measurement for Harsh Environments


Hello Tomorrow

Dubai inspires new connections

emirates.com/usaSelect 68 at www.HydrocarbonProcessing.com/RS

Linde Process Plants, Inc.6100 South Yale Avenue, Suite 1200, Tulsa, Oklahoma 74136, USAPhone: +1.918.477.1200, Fax: +1.918.477.1100, www.LPPUSA.com, e-mail: [email protected]

A member of The Linde Group

StarLNGTM

The leading small-to-mid-scale standard LNG plant

Linde Process Plants, Inc., a world leader in cryogenic technologies, has now translated the idea of product standardization into the LNG industry. These modularized plants cover about 90% of real life boundary conditions, and provide improved, simple and robust technology with high

StarLNG™ delivers a pre-engineered process design, standard documentation and modularized plant layout for shortest delivery time with minimum on-site construction.

Consider Linde Process Plants, Inc. for your next project.Select 85 at www.HydrocarbonProcessing.com/RS

Impact

�15

tor demand will rise at a faster pace. Or-ganic inhibitors will be the primary ben-eficiaries as companies look to develop new products that offer more protection at lower treat rates, and that are more cost-effective than existing alternatives. Replacing molybdates, where possible, will remain a top priority as molybdate prices remain comparatively high, and some concerns about their environmen-tal impact have arisen.

Poll says US citizens want more solar, wind and natural gas

No fewer than two in three US citi-zens want the country to put more em-phasis on producing domestic energy using solar power (76%), wind (71%) and natural gas (65%). Far fewer want to emphasize the production of oil (46%) and the use of nuclear power (37%). Least favored is coal, with about one in three respondents wanting to prioritize its domestic production. These are the results of an opinion poll administered by Gallup, based on telephone interviews conducted March 7–10, with a random sample of 1,022 adults, aged 18 and old-er, living in all 50 US states and the Dis-trict of Columbia.

Democrats’ and independents’ top choice is solar power, while natural gas places first among Republicans. Republi-cans and Democrats disagree most on the priority that should be given to oil as a fu-ture energy source, with 71% of Republi-cans wanting more emphasis placed on it, compared with 29% among Democrats. Republicans are also much more sup-portive than Democrats of coal (51% vs. 21%) and nuclear power (49% vs. 30%).

Where people live in the US makes a difference in their views about which sources of domestic energy they want the country to emphasize more. People in the South tend to be more supportive of traditional energy sources such as oil and coal than are those in other regions.

Still, for respondents in every region, including the South, solar power is the top choice, or is tied for the top spot, among the energy sources tested. The US has a great opportunity to accelerate its economic growth over the next sev-eral years by emphasizing and using its enormous energy resources to produce domestic energy. But there has been no

consensus among US citizens about how to optimize domestic energy production while preserving the environment.

US citizens overall and across political and socioeconomic groups generally are most likely to call for more emphasis on solar and wind power, but these potential future sources of energy have a long way to go in terms of technology and affordability before they can significantly affect overall US domestic energy production. Ameri-

cans are also sharply divided politically over achieving greater domestic energy production using more traditional energy sources such as oil, coal and nuclear power.

This leaves natural gas, which 59% of Democrats, 62% of independents and 79% of Republicans say should have more emphasis in the US. The technology exists and is being implemented to allow natural gas to become a more significant contribu-tor to US domestic energy production.


ExcellentMineralsSolutions

Expertise where it counts.SM

8625 Grant Rd.St. Louis, MO 63123T: +1 314 843-4437F: +1 314 843-7964Email: [email protected]

LEWIS® PUMPSVertical Chemical Pumps

All pathslead to

Lewis® Pumps

Lewis® pumps are the world standard for pumps and valves in the sulphur chemicals industry. Offering a family of steam-jacketed sulphur pumps, outstanding reliability in high-temperature sulphuric acid, and new designs for molten salt energy transfer, Lewis continues its long tradition of superior products and services.

Standard replacement parts are always available. Emergency service or parts to any major airport worldwide within 72 hours. No matter what your application, when you need a superior product with exceptional service... all paths lead to Lewis. Customers in over 100 countries can’t be wrong.



Associations

ADRIENNE BLUME, MANAGING EDITOR

[email protected]

EMGC 2013 explores Eastern Mediterranean resource potential

Gulf Publishing Company held its in-augural Eastern Mediterranean Gas Con-ference (EMGC) at the Hilton Cyprus in Nicosia, Cyprus, from April 8–10. The event surpassed expectations in terms of both attendance and interest, and gave ex-ecutives from operating, service and tech-nology companies the chance to share in-sight into development opportunities for this exciting resource area.

The conference featured speakers from top companies operating in the re-gion, including Noble Energy Inc., Hy-perion Systems Engineering Group, Cy-prus National Hydrocarbons Co., Total E&P Research and Technology USA LLC and others.

Monday workshop. The conference opened with the “Doing Business in the Mediterranean” workshop on Monday, and continued with two days of techni-cal presentations, networking events and a gala dinner on Tuesday. The workshop, hosted by EMGC media sponsor De-loitte, drew over two dozen attendees seeking to share knowledge and gain in-formation about economic conditions and tax and customs laws in Cyprus and elsewhere in the region.

Speaker Paul Mallis, Cyprus Oil and Gas Tax Leader for Deloitte, noted that Cyprus’ low corporate tax rate of 10% (which will move to 12.5% after the re-cent, €10 billion agreement with the Troika), is an attractive reason for doing business in Cyprus.

Day 1: Offshore, LNG and national opportunities. EMGC continued on Tuesday with the first day of technical presentations. Executive speakers focused on the regional implications and potential of the Eastern Med’s new energy resourc-es. Delegates from the Israeli and Cypriot governments, along with industry execu-tives, examined offshore oil and gas op-

portunities and challenges for Cyprus, Israel and the EU as a whole.

The LNG question. Several speakers expressed the need for up to three lique-fied natural gas (LNG) export trains, with a total capacity of 15 million tons per year, to deliver Eastern Med gas to other Euro-pean nations. Cyprus’ Deputy Director of Energy Services for the Ministry of Indus-try, Commerce and Tourism, Constanti-nos Xichilos, noted that Noble Energy Inc. and partner Woodside Petroleum Ltd. are studying a possible LNG terminal site at Vasilikos in Cyprus.

Noble Energy’s take. Noble Energy Chairman and CEO Charles Davidson spoke at length on Tuesday morning about the discoveries his company has made in the region, and the implica-tions these resources have for the Eastern Med’s energy future.

“The region is evolving into a major energy player in the world, not only for countries in the region, but also for the countries they will be working with,” said Mr. Davidson. The massive gas resources available in the Eastern Med will provide enough clean, low-cost energy to displace all of the fuel used by automobiles in Is-rael for 14 years, Mr. Davidson noted.

“We have to work cooperatively with all of our partners, especially [the region-al] governments,” to make these projects work, Mr. Davidson said.

The Israeli perspective. AJM De-loitte Partner for Energy and Resource Advisory Services, Robin Mann, noted that Israel could be self-sufficient in gas by 2016. Gas from the Tamar and Le-viathan fields will supply power for the country’s electricity needs, and excess gas from these fields could be sent as LNG to Asia or Europe. Alternatively, the gas could be piped through an undersea pipe-line through Cyprus or Turkey and then onto Europe. Another option would be to set up a floating LNG (FLNG) facility to process and transport the gas.

Emerging markets. Tuesday after-noon featured presentations on the mar-ket for the new resources, with executives

from DNV, KBR Inc., and GE Oil and Gas Turbomachinery offering perspectives on natural gas mega-projects, including pro-posed LNG and FLNG projects.

Concluding the afternoon session was a panel discussion on the impact of the new energy resources, featuring panelists Rony Halman, founder and Chairman of Israel Opportunity Oil & Gas Co.; Wafik Bey-doun, President and CEO of Total E&P Research and Technology USA LLC; and Philip Hagyard, Senior Vice President of Gas Monetization at Technip.

Tuesday evening gala. Day 1 con-cluded with a heavily-attended gala din-ner at the Hilton Cyprus sponsored by Deloitte Ltd. The dinner featured a key-note speech by Noble Energy’s Charles Davidson (FIG. 1), who said that the East-ern Med’s gas resources would soon be moving to the export stage, which would inevitably involve energy exchange with Middle Eastern nations.

Mr. Davidson emphasized that the work being done in the Eastern Med by Noble Energy and its partners is a joint effort. “These aren’t only multi-billion-dollar projects; they’re multi-decade projects,” Mr. Davidson said. For such projects to be successful, stable tax en-vironment and investment climates are needed, along with a streamlined infra-structure process.

FIG. 1. Noble Energy Chairman and CEO Charles Davidson addressed attendees at Tuesday evening’s gala dinner.

Sure we can do the standard tower open/clean/inspect/close work but it’s those tough and challenging jobs that have helped us earn our stripes.

We recently completed a revamp on one of the largest vacuum towers in the western hemisphere, much to our customer’s satisfaction. We’ve mastered a resection method that is an excellent and cost saving alternative when footprints are tight.

Not long ago we tackled a 114-vessel project at a natural gas processing plant. Our team blinded, opened, cleaned, inspected and repaired all of the vessels in just 14,642 man-hours; less time than had been scheduled and well under budget.

When performing any number of services, we don’t overlook anything from external pipe fl anges, complicated vessel internals, feed/draw arrangements, section replacements, nozzle and strip lining installation/repair. We strive for “Zero Injury.” As for quality, well, that’s why our customers invite us back again and again and give us their annual maintenance/service contracts.

We thrive on accepting challenges then exceeding expectations but we are just as agile with a single tower project as we are with plant-wide turnarounds. We are quick on our feet and can mobilize swiftly for emergencies.

TIGER TOWER SERVICES 3012 Farrell Road, Houston, TX 77073

Phone: 281.951.2500 • Fax: 281.951.2520

www.tigertowerserivces.com

Now Accepting Challenging Projects

Challenge us today and we’ll have your towers productive tomorrow.


Mr. Davidson concluded his speech by saying, “Our goal is to build on what we’ve done already, and we hope we can look back a few years from now and say that we created a forum for discussion [with this conference] for what is happen-ing in this region.”

Day 2: Infrastructure, resource devel-opment and beyond. The second day of EMGC’s technical program addressed re-gional infrastructure developments in the morning, while the afternoon focused on resource development and the future of the Eastern Mediterranean.

FLNG outlook. Victor Alessandrini, FLNG Business Development Manager for Technip’s Offshore Business Unit, spoke about the market and future devel-opments for FLNG (FIG. 2). The tech-nology is a cost-optimized, mobile and environmentally non-invasive way of pro-cessing gas into LNG and transporting it to various locations, Mr. Alessandrini said.

Some technical challenges of FLNG include incorporating a full LNG plant onto a floating production, storage and offloading (FPSO) unit; gas processing facilities must be adapted to a marine environment. Space, weight and stability management are other important consid-erations when planning an FLNG project, Mr. Alessandrini noted.

Cypriot growth aspirations. Dr. Symeon Kassianides, Chairman and CEO of Hyperion Systems Engineering Group, next examined Cyprus’ readiness to ad-dress the challenges and opportunities be-fore it. Dr. Kassianides noted that both the public and private sectors are preparing the country for an upswing in work oppor-tunities in the oil and gas sectors. He also confirmed that FEED studies are being performed for an LNG terminal in Cyprus.

According to Dr. Kassianides, long-term potential projects in Cyprus include a methanol plant, an ethylene facility, a possible gas-to-liquids (GTL) plant and other facilities. “We strongly believe in Cyprus as an upcoming energy hub for the future of the Eastern Mediterranean,” Dr. Kassianides said.

Israel’s energy goals. The Wednesday afternoon session kicked off with a presen-tation by Jay Epstein, Business Develop-ment Manager for Israel Natural Gas Lines, on the infrastructure needed for the devel-opment of Israel’s natural gas sector. Af-terward, Chairman of Israel’s Dor Chemi-

cals Ltd., Gil Dankner, spoke to attendees about the use of natural gas and its deriva-tives as alternative transportation fuels.

The future of the Eastern Med. During EMGC’s eighth and final session, the CEO of Cyprus National Hydrocarbons Co., Dr. Charles Ellinas, noted that the establish-ment of an LNG terminal is vital, as it will enable Cyprus to access markets in Europe, the Far East and other global regions.

Concluding the session was Noble En-ergy’s Director of Operations for the East-ern Mediterranean, Terry Gerhart. Mr. Ger-hart detailed Noble Energy’s development plans for the Tamar, Leviathan and Cyprus A fields. Additionally, an LNG plant to pro-cess this gas could begin operations near the end of the decade and accommodate up to three trains, Mr. Gerhart said.

Inaugural EMGC a success. During his closing remarks on Wednesday afternoon, Gulf Publishing Company President and CEO John Royall thanked attendees for helping make the inaugural EMGC a suc-cess. Mr. Royall said that the conference’s networking, forum-building and aware-ness-raising objectives had exceeded orig-inal expectations.

Major companies such as Noble Energy, Eni, Total, Woodside Petroleum and oth-ers have made significant investments in the region, Mr. Royall noted. “My advice to the industry and to all of you in this room is to look where the experts are investing their money,” he said. “And they’re invest-ing it here, in the Eastern Mediterranean.”

Gulf Publishing Company plans to hold its second annual Eastern Mediter-ranean Gas Conference in Tel Aviv, Israel, in 2014.

Associations

www.borsig.deBORSIG GmbHPhone: ++49 (30) 4301-01Fax: ++49 (30) 4301-2236 E-mail: [email protected]

Egellsstrasse 21, D-13507Berlin/Germany

SERVICES

for

Pressure vessels and

heat exchangers

Power plants

Mechanical Engineering

Plant and pipe con-

struction

Ball valves and valves

Compressors

Membrane Technology

BORSIG - Always your first

choice

BORSIG


FIG. 2. Technip’s Victor Alessandrini discussed the global outlook for FLNG developments on Wednesday morning.

l l l



Reliability HEINZ P. BLOCH, RELIABILITY/EQUIPMENT EDITOR

[email protected]

Manage your time constructively

We often view ourselves as very effi-cient reliability implementers. Yet, as pro-fessional engineers, we take action to save time, accomplish more and reduce stress by applying better time-management skills. Being productive requires effort. Likewise, resourcefulness is needed to improve individual performance.

More important, we must do a better job of balancing our personal and profes-sional lives; rest and relaxation from the harried duties from work is a positive. In all of our professional endeavors, we should strive to be better contributors in-stead of mere consumers.

Set realistic goals. Serious contribu-tors often set daily goals. Why? Goals are important. Meeting daily goals pro-vides a sense of accomplishment and an opportunity to add value to our compa-nies. Value-adders set and readjust their priorities every day. Many tasks require concentration and organization. Are your tools close at hand? A hammer is a tool, but so is a micrometer, a checklist or a good technical text.

Plan your day. If phone calls are neces-sary, strive to make them when you have the best chance of reaching your desired person. E-mailing your contact with a brief and complete message is a more time-effi-cient action. Likewise, send e-mails with written summaries to the appropriate dis-tribution list for action.

Meetings. In planning meetings, keep them brief and stay on schedule. To im-prove everyone’s productivity, start on time and finish on time, and stick to the agenda. When attending a meeting, never miss out on the opportunity to say nothing. Repeat-ing what has already been said, or what can be confirmed in a written summary, takes away valuable time from others.

Delegate when possible. Delegating tasks can accomplish more with the pres-ent staff and provide training opportuni-ties. Also, it conveys to the team members

that they are valued. Ask for “buy-in” from those to whom you delegate a task, and firmly determine a mutually acceptable delivery schedule.

Beware of paper shuffling. Productive professionals strive to handle a piece of pa-per only once. They resist the temptation of moving papers among temporary piles. For example, a freelance writer once spent eight weeks massaging a four-page article. The freelancer’s staccato work pace ad-versely affected the efficiency of others.

Divide and conquer onerous tasks. Some days appear to be an endless list of tasks that add more dread to your job and attitude. One solution is to compile a complete list of work and project items to be handled; put everything on the writ-ten list. As new tasks materialize or are assigned, add them to the list. Visually crossing off completed items or tasks def-initely improves your mood and attitude as the “things to be done” are completed. Many professionals keep the “to do” list visually close by, such as on a computer screen. It is easier to stay on schedule with such queues refocusing the individual.

Set realistic priorities. Priorities are of-ten an issue. Let your boss assist you in set-ting and resetting priorities. Break down large tasks into small segments; this will re-move intimidation from the total endeav-or, and provide attainable and achievable goals. Be realistic in timing task durations.

Of course, you can assign priorities to the tasks and activities according to importance. It is difficult to clearly dis-tinguish between “urgent” and “impor-tant.” There may even be some off time within your day. This is the opportunity for reliability professionals to broaden their knowledge base. They can work on tool making, repair tasks or specification updates. It is an opportunity to develop technical articles and reports that will add value to others, or to scan some technical articles for later use.

Never procrastinate. Remember to manage your activities via their priori-ties on your task list. Stay in control and keep bosses informed of schedule chang-es. Task lists should be flexible. The ob-jective is to maintain control so that the tasks accomplished each day are done by choice, and not by chance.

Maintain focus. Seasoned pros do not rush from job to job or worry about doing everything that they have listed. Time-management consultant Alan Lakein stresses that one rarely reaches the bottom of a “to do” list. He remindes his students that it’s not completing the list that counts, but making the best use of one’s time. We should strive to accomplish the bulk of tasks that are truly important and warrant our skills. When possible, delegate the unfinished items to others or transfer the project to tomorrow’s list. Ask if finishing the job produces significant benefits. If not, it may not be a high-priority task.

What is ‘urgent’ or ‘important?’ Some tasks yield better results than others. When looking over a list of duties, consider the results that each one will bring. True, at first glance, everything on the list seems urgent. Still, we should ask if urgent matters are al-ways important, deserving a major time in-vestment. Michael LeBoeuf, a professor of time management at the University of New Orleans, makes this observation: “Impor-tant things are seldom urgent, and urgent things are seldom important. The urgency of fixing a flat tire when you are late for an appointment is much greater than re-membering to pay your auto insurance premium, but its importance [the tire] is, in most cases, relatively small.” Then he la-ments: “Unfortunately, many of us spend our lives fighting fires under the tyranny of the urgent. The result is that we ignore the less urgent but more important things in life. It’s a great effectiveness killer.”

It is more rewarding to work at some-thing that yields important results than it is simply to be busy at whatever activity is

Reliability

22

at hand. Try to direct your efforts to activ-ities that result in true accomplishments.

80/20 rule. A number of time-manage-ment experts believe that we can narrow the top-priority items down to about 20%. These experts cite, as a guide, the 80/20 rule. This principle was formulated by the 19th-century Italian economist Vilfredo Pareto; it states that only about 20% of the causes produce about 80% of the results.

But how can the 80/20 rule be applied to your time management? Analyze the items on your “to do” list. Perhaps, you can be 80% more effective by accomplish-ing 2 out of the 10 items listed. If so, those two items are important on your list. Also, analyze a project before diving in. How much of it is truly important to your ob-jective? What part of the job will produce the most significant results? This portion of the task is a priority.

Time-management consultant Dru Scott, after discussing Pareto’s principle, explains how to make it work for you. She says: “Identify the vital ingredients neces-sary to achieve your objective. Do these things first. You will get the most results in the least amount of time.”

Example. One refinery’s statistics from the 1970s established that 7% of its 3,200 process pumps experienced a disproportionate share of pump outage events. Slightly over 60% of the money spent on pump maintenance was allo-cated on a chronic 7% pump population. These pumps were quite obviously prob-lematic machines; they received more at-tention than the average refinery pumps.

Applying the discussed principles, what percentage of your day’s activities would you expect to categorize as top pri-ority? Of course, that will depend upon your specific responsibilities.

Enjoy the benefits. Now, we can appreci-ate that being the master of your time is not a matter of being preoccupied with never wasting a minute or rushing from crisis to crisis. Rather, effective time management means selecting the appropriate task for the present conditions. It means discern-ing what activities yield the best results and then spending time on those tasks.

There are no fixed rules for personal organization. Remember: Be flexible, ex-periment and adapt. Discover what works best for you. By gaining better control of your time, you will find more sense of ac-complishment each day. Although more will remain for tomorrow, there will be satisfaction in directing your efforts to the most important things. There is enough time to complete the tasks that matter. Do not be a victim of hectic circumstances; be the master of your time.

HEINZ P. BLOCH resides in Westminster, Colorado. His professional career began in 1962 and included long-term assignments as Exxon Chemical’s regional machinery specialist for the US. He has authored over 500 publications, among them 18 comprehensive books on practical machinery management,

failure analysis, failure avoidance, compressors, steam turbines, pumps, oil-mist lubrication and practical lubrication for industry. Mr. Bloch holds BS and MS degrees in mechanical engineering. He is an ASME Life Fellow and maintains registration as a Professional Engineer in New Jersey and Texas.


Team is a world-class service company with the right people, technology and experience needed to keep your plants online

and in production. Our highly skilled technicians work to earn your continued trust and conf idence one job at a time.

| www.teamindustrialservices.com

Scheduled maintenance, inspections, emergency response…Team delivers

LEAK REPAIRS

FIELD HEATTREATING

FIELD MACHINING

VALVEINSERTION

HOT TAPS /LINE STOPS

TECHNICAL BOLTING

VALVE REPAIR

NDE/NDT INSPECTION

EMISSIONS CONTROL

PIPE REPAIRSERVICES

TURNAROUNDSERVICES

PIPELINE SERVICES


relentless

UOP has been serving catalyst customers for decades, and constant innovation ensures we’ll be there for decades to come.

Pilot plants. A true commitment to fast, thorough research and

development. A vast team of experienced, dedicated engineers

and scientists. The reasons to choose UOP catalysts are many,

but perhaps the biggest of all is that UOP never quits innovating. When you use UOP

catalysts, you’ll know you’re using products specifically designed to help optimize

your operation. UOP invented the refining and petrochemical technology used in most

operations today, and that knowledge and expertise comes through in every catalyst we

develop. You’ll always get the high stability and global support your business needs, and

you’ll be working with a company that will keep you on the leading edge year after year.

For more information about UOP catalysts, visit www.uop.com/catalysts.© 2013 Honeywell International, Inc. All rights reserved


Integration Strategies

RALPH RIO, CONTRIBUTING EDITOR

[email protected]

Asset information management improves life cycle benefits for equipment

Traditionally, the various production management and asset management systems in the hydrocarbon processing industry (HPI) and other heavy industries resided in silos with separate management, architecture and technologies. Asset performance management (APM) involves collaboration among production and maintenance groups to improve execution with higher pro-ductivity, reduced risk and improved return on assets (ROA).

The asset and production management systems have large amounts of data and information that can be used for optimi-zation across these systems. Result: Improved performance for both maintenance and operations. Asset information management (AIM) provides the means for optimization across the APM spectrum. AIM for APM offers new oppor-tunities for achieving key performance indicators (KPIs) and C-suite objectives.

APM and equipment life cycles. APM applies to the long-term, “operate and maintain” phase of an asset’s life cycle, which defines revenue, margin and profits. An APM strategy involves integrating production management (making the product) with maintenance (ensuring the capability to produce). This enables both production and maintenance management to align and meet objectives with higher productivity, reduced risk and improved ROA.

APM includes sharing asset management information with collaborative production management (CPM) or manufac-turing execution system (MES) applications. Such collabo-ration provides visibility for new opportunities to improve asset availability. It includes an enhanced understanding of risk, with fact-based risk assessment. Also, APM enables or-ganizations to uncover opportunities to balance operational constraints and improve ROA.

A range of applications do come together in the APM do-main. Some have traditionally been associated with asset man-agement, including enterprise asset management (EAM), predictive maintenance (PdM), reliability and condition moni-toring. Others come from production management, including MES, CPM, quality management, laboratory-information man-agement systems, and historians.

AIM for APM. Good information management leverages smart assets for better operations. Smarter assets include remote di-agnostics, service contracts for suppliers of complex equip-ment and higher uptime for the users. Smarter operations aids improved production scheduling (including maintenance), yield/quality (via value-added performance) and asset longev-ity (reducing equipment stresses). Together, smarter assets and

operations provide opportunities for managing the entire plant including maintenance costs and energy management.

Asset management systems require information about the plant equipment to function. Managing this information is criti-cal. This asset (equipment) information includes data about the asset over its entire life cycle.

Managing asset information involves the business processes and technology for:

• Organizing the structure to classify the information for consistent, accurate record-keeping and extraction for analysis and reporting in a robust, consistent manner

• Creating information entry and storage• Controlling access to information according to role-spe-

cific user needs and authority• Utilizing change of management to modify information

with proper authorization and timing, which includes recording additions, changes and deletions; it is an audit trail

• Auditing the asset-related data.

AIM information types. AIM information pertains to both structured and unstructured data:

Documents contain reference materials including drawings and standard operating procedures. Examples include draw-

19%

22%

29%

43%

46%

51%

54%

57%

67%

44%

40%

28%

37%

43%

34%

34%

35%

25%

63%

62%

57%

80%

89%

86%

88%

93%

92%

0% 20% 40% 60% 80% 100%

Project cost capitalization

Capital project plan and execute

Limited previous software

Safety and risk management

Visibility to management

Improve quality or yield

Maintenance cost control

Extend asset longevity

Improve uptime (MTBF, MTTR)

HighMediumTotal

Source: ARC survey with 134 respondents January 2013.

FIG. 1. Business drivers for EAM.


Integration Strategies

ings from the design-and-build phase of the asset life cycle or upgrades during the operate-and-maintain phase.

Transactions have a predetermined set of data fields, which are transferred concurrently to provide an activity record. For asset management, examples include work orders for mainte-nance and inspections in the EAM system.

Process data includes real-time process values and time-stamped recorded values. This data provides the basis for scheduling PdM. It also provides the input for making reports.

Reports, both online and offline analysis, can include inci-dent reports, KPIs and analysis for assessments of problems or improvements.

Usually, each organization manages its own information. This leads to independent technology, applications and structures. Inconsistencies in organizing, creating, controlling, change man-agement and asset hierarchies inhibit collaboration and under-

mine information sharing information. These inconsistencies all weaken an APM strategy.

Why change for APM? ARC’s annual survey of end users re-garding maintenance management systems consistently yields high ranks for these key metrics: uptime, asset longevity, cost con-trol, quality/yield and safety. With the increasing speed of busi-ness, visibility into current status has grown in importance (FIG. 1). This requires that asset information is managed with role-based visibility such that only the most current and pertinent information is visible to each user. Achieving and exceeding these metrics require optimization across the various systems for maintenance and operations.

These goals relate directly with C-suite metrics, which are in the P&L statement and balance sheet (TABLE 1). Executes respond to their metrics much like you do. When you have a positive effect on these metrics, you get and retain C-suite attention. This helps with obtaining the resources needed to be successful.

PanaFlow HT

In hazardous environments safety comes first

SILCer tified

GE

RALPH RIO, research director for Enterprise Systems, has been with ARC since 2000. Prior to ARC, he served in various marketing management roles at GE Fanuc Automation (GEF), Intellution, Digital Equipment Corporation, and Codex Corporation, as well as manufacturing engineering roles at Texas Instruments and General Electric. He holds a BS degree in mechanical engineering and an MS degree in management science from Rensselaer Polytechnic Institute in Troy, New York.

TABLE 1. Comparing key maintenance and executive KPIs

EAM KPIs Financial reported aff ected C-suite KPIs

Uptime P&L and balance sheet Revenue increase and less inventory

Asset longevity Balance sheet Cash conservation

Cost control P&L Profi tability

Safety Annual report Risk management


©2013 John Zink Company LLC. JOHN ZINK HAMWORTHY COMBUSTION is a trademark of John Zink Company LLC. JZ, KEU, KALDAIR and TODD are registered trademarks of John Zink Company LLC. HAMWORTHY COMBUSTION is a registered trademark of Hamworthy Combustion Engineering Ltd. COEN is a registered trademark of Coen Company, Inc. AIROIL-FLAREGAS is a trademark of Hamworthy Combustion Engineering Ltd. HAMWORTHY PEABODY is a trademark of Coen Company, Inc.

Emissions+, the FREE emissions and conversions calculator by John Zink Hamworthy Combustion. Available in the iTunes App store.

International Headquarters11920 East Apache StreetTulsa, Oklahoma 74116United States

T: +1.918.234.1800 F: +1.918.234.2700 johnzinkhamworthy.com



Boxscore Construction Analysis

LEE NICHOLS, DIRECTOR, DATA DIVISION

[email protected]

Jurong Island—Asia’s preemptive LNG trading hub?

In the South Pacific, the artificial is-land of Jurong Island is the heart of Sin-gapore’s energy and chemicals industry. It houses over 100 leading global petroleum, petrochemicals and specialty chemicals companies. With a rich history as an oil trading hub, Singapore is on course to become a regional liquefied natural gas (LNG) trading center with the construc-tion of its first LNG terminal.

Developed by Singapore LNG, the Singapore LNG terminal is being con-structed on a 30-hectare plot at the Meranti Seafront on Jurong Island. The $1.7-billion (B) facility will be used for importing LNG, reloading and regasify-ing, and storage. It is the first terminal constructed with bidirectional opera-tions and storage. The new terminal will allow companies to unload LNG car-goes, store them, and then ship them at a later date. It will also aid in the redistri-bution of LNG supplies to regional des-tinations that cannot build large import facilities or that have ports too small to handle mega-sized LNG vessels.

The original contract to construct and operate the terminal was assigned to PowerGas in 2007. In 2009, Singapore’s Energy Market Authority (EMA) took over control of the project and created Singapore LNG (SLNG) to continue de-velopment. In 2010, SLNG awarded the engineering, procurement and construc-tion (EPC) contract to Samsung C&T. TABLE 1 lists additional awards.

The first phase is scheduled to be completed by mid-2013. This phase in-volves the construction of one jetty and two storage tanks, each with a capacity of 188,000 cubic meters (m3) and the abil-ity to process 3.5 million tons per year (MMtpy) of LNG. The primary jetty (FIG. 1) is being constructed to handle the newest and largest Q-max vessels and LNG carriers. A third, 188,000-m3 stor-age tank and two additional jetties will be commissioned in 2014, raising total storage capacity to 6 MMtpy. A fourth

tank is scheduled to come online in 2017 at a cost of $500 MM, raising site capac-ity to 9 MMtpy. The SLNG terminal’s master plan provides for seven storage tanks and a peak capacity of 20 MMtpy.

With LNG demand increasing in Asia, the government of Singapore is contemplating the construction of a sec-ond LNG terminal. In November 2012, the EMA issued a tender for a consultant to conduct a six-month feasibility study. The second terminal could double Sin-gapore’s existing storage capacity.

SLNG received the first LNG cargo from Qatar Operating Co. Ltd. (Qatar-gas) in the first quarter of 2013. This load will be used to commission the SLNG ter-minal before full operations begin in the second quarter. Singapore has confirmed BG Singapore Gas Marketing (BGSGM), a BG Group subsidiary company, as the LNG aggregator for train one. BG’s ex-clusive license allows it to import LNG and sell regasified LNG in Singapore up to 3 MMtpy, or until the year 2023. BG will supply the terminal from its exten-

TABLE 1. Project awards for the SLNG terminal

GDF Suez and PowerGas Front-end engineering and design (FEED) study.

Samsung C&T EPC, awarded February 2010. Also awarded contract for third storage tank and Secondary Berth Project.

Fluor Engineering and related management services.

Foster Wheeler Asia Pacifi c Project management consultancy services.

WorleyParsonsBasis for design and FEED design. Development of EPC contract tender documents along with subsequent assessment and recommendations.

FIG. 1. View of the Singapore LNG terminal’s primary jetty. Photo courtesy of Singapore LNG Corp.

30

Boxscore

sive LNG portfolio and supply positions in Trinidad, Egypt, Nigeria, Equatorial Guinea and the proposed Queensland Curtis LNG facility in Australia. This baseload demand will start the LNG project with additional supplies coming from a combination of LNG and piped natural gas, short-term and long-term contracts, and spot cargoes.

BGSGM has already signed over 1.5 MMtpy of gas sales contracts with a vari-ety of customers in Singapore. The com-mitments include six large-scale power- generation companies including Senoko Energy, PowerSeraya, Tuas Power Gen-eration, SembCorp Cogen, Keppel Mer-limau and Island Power Co.

The additional gas supplies by SLNG will aide in domestic use. Gas-fired plants produce 80% of Singapore’s electricity needs. This percentage is expected to in-crease to 90% within the next few years. Unable to fulfill demand, Singapore has been purchasing gas supplies from Indo-nesia and Malaysia. The SLNG terminal should cover the gap between supply/demand and lessen natural gas supplies via pipelines from neighboring countries.

Benefits. The International Energy Agency (IEA) forecasts Asian LNG con-sumption to annually increase to 790 Bm3

by 2015. Within that time, Southeast Asia is scheduled to construct a dozen import terminals resulting in over 35 Bm3 of ad-ditional capacity. At present, no real LNG trading hub has been established within the region. It is difficult for companies to purchase LNG at larger (cheaper) quan-tities. Imports are linked to the price of oil; these prices are five times higher than those in North America, which are based on Henry Hub pricing. An Asian LNG trading hub would allow pricing to better reflect supply and demand fundamentals. Singapore’s strategic location, business-friendly environment and bidirectional LNG terminal and storage facilities make it a prime candidate to become Asia’s LNG trading hub.

Location. Singapore’s greatest benefit is its central location to the Asian market. It is nestled between LNG demand centers in Northeast Asia and LNG supply sourc-es in Southeast Asia, the Middle East and Australia. It also sits at the southern tip of the Strait of Malacca, one of the most im-portant shipping lanes in the world and a

REEal improvement of your compressors

www.hoerbiger.com

For more information: [email protected]

REE Rel

ia

ble Effi cient

Environmentally

soun

d10

9

8

7

6

5

4

3

2

1

0

Upgrade & Revamp

REEal improvement of your compressors

The continuous improvement of compressors

is fundamental if plants are to meet modern

demands in terms of run cycle, energy

efficiency and legal compliance. In order to

do this, clear and accurate data on the real

performance of your compressor is required.

The REE Assessment carried out on-site by

HOERBIGER senior compressor experts

assesses your equipment using the REE index.

It allows you to see where there is room for

improvement and where savings can be made,

helping you to choose the right technology and

make the right business decisions.

Ever worried about the reliability, energy efficiency and

environmental soundness of your compressors? Not any more!



�31

gateway to Southeast Asia. This strategic location allows LNG traders to centrally store their LNG cargoes and ship them to heavy consumers such as China, Japan and South Korea.

Singapore’s business-friendly environ-ment is also a major selling point. The government has enacted various incentive programs and tax breaks to attract major oil and gas companies. The Global Trader Programme (GTP) was instituted in 2001 to encourage companies to establish re-gional/global operations headquarters in Singapore. Companies can benefit from concessionary tax rates of 10% on their qualifying trade income. In 2007, the government adopted a 5% concessionary corporate tax rate for LNG trading in-come with the intent to spur the develop-ment of a LNG trading hub.

In response to these benefits, LNG companies have flocked to Singapore. Within the last five years, the number of major LNG players has gone from zero to 14; they include BG, BP, ConocoPhil-lips, Gazprom, GAIL, GDF Suez, Shell and Statoil.

Challenges. Although Singapore bene-fits from its central location to LNG-hun-gry Asian markets, a corporate-friendly atmosphere and a new bidirectional LNG terminal facility, it still faces regional and global challenges before it can emerge as Asia’s LNG trading hub.

Singapore is in stiff competition with other Asian LNG players that are actively trying to establish regional LNG hubs. The stiffest competition stems from Ma-laysia; this nation is constructing its own bidirectional LNG terminal at Pengerang in Southern Johor. The first construc-tion phase of the $1.3-B Pengerang In-dependent Deepwater Petroleum Ter-minal (PIDPT) is set for completion in 1Q 2014. Located within the Pengerang Integrated Petroleum Complex (PIPC), PIDPT is being developed by the Johor state government, Netherland’s Royal Vopak and Malaysia’s Dialog Group. It will have a total storage capacity of 5 MMm3 and allow for storage, loading and regasification of LNG for trading and domestic use. PIDPT’s construction will coincide with Petronas’ Refinery and Petrochemical Integrated Develop-ment Project (RAPID). RAPID is a $20-B integrated refinery and petrochemical complex located at the PIPC. The proj-

ect will consist of a 300,000-bpd refinery capable of producing 9 MMtpy of petro-leum products, and the state-of-the-art petrochemical complex will produce 4.5 MMtpy of downstream petrochemicals. The RAPID project aims to transform Southern Johor into a new Asian petro-chemical hub.

Petronas is also constructing a ninth LNG production train at its Bintulu com-plex. The new LNG train will add 3.6 MMtpy of capacity. Once completed, the Petronas LNG Complex will have a combined capacity of over 27 MMtpy—creating one of the world’s largest LNG production facilities at a single location.

Malaysia is also set to commission the world’s first LNG regasification unit lo-cated on an island jetty. The 3.8-MMtpy Melaka terminal is located approximate-ly 3 km offshore Sungai Udang Port. Two floating storage units (FSUs) will receive and store LNG; the facility has subsea and onshore pipelines connected to the Peninsular Gas Utilization pipeline net-work. Melaka is expected to receive its first LNG cargo load in August 2013.

South Korea, Japan and Indonesia are also possible Asian LNG hubs. Large en-ergy-consumer countries—South Korea and Japan—already have the necessary facilities and storage tanks to trade LNG but lack the centralized location. Indone-sia is striving to increase LNG export ca-pacity with BP’s Tangguh LNG expansion project and Mitsubishi Corp., Pertamina, and Medoc’s joint venture Donggi-Se-noro LNG project. Other Asian countries are embracing the use of floating liquefied natural gas (FLNG) and floating storage and regasification (FSRU) units in lieu of pricey onshore terminals capable of han-dling large LNG vessels.

Singapore’s plans are challenged by the recent development of US LNG exports. With the Panama Canal expansion to be completed in 2015, it is unknown if US LNG exports will be routed through an LNG hub like Singapore or sent directly to Asian consumers such as Japan, Thai-land, the Philippines and South Korea. The Government of Singapore Invest-ment Corp. has even made an investment in Cheniere’s Sabine Pass LNG export

Advanced Inspection and Assessment Solutions for Process Fired Heater Coils

Learn more about FTIS™ through animation on our website:www.QuestIntegrity.com/FTIS

1200 Process Heaters in 35 Countries Since 1997

+1 253 893 7070

A TEAM Industrial Services Company

Improve heater reliability Minimize heater downtime




terminal to take advantage of cheap US gas supplies. Whether or not that gas will be sent to Singapore is unknown.

An Asian LNG hub could reduce nat-ural gas prices dramatically throughout this region. Singapore’s combination of geographic, economic and infrastructure advantages will facilitate its efforts as the preemptive Asian LNG hub. However, competition from other Asian countries, primarily Malaysia, could curtail these plans. Regardless, Singapore continues to attract significant LNG players eager to make large investments in its emerg-ing LNG sector.

New Database. Hydrocarbon Process-ing will unveil the new, enhanced Con-struction Boxscore Database in May. For more than 60 years, the Construction Boxscore Database has provided oil and gas professionals with real-time informa-

tion on refining, petrochemical and gas-processing construction projects from around the globe.

Boxscore has deep roots in the hydro-carbon processing industry since its in-augural publication in Petroleum Refiner, the forerunner to HP, in August 1947. In the 1950s, “Box score,” as it was called, was one page with a list of 80 projects. The latest-generation Boxscore contains over 3,500 projects in over 125 countries. New, enhanced search functions allow us-ers to search broadly or pinpoint specific projects by location. The new project data page provides users with detailed project information on the operating, engineer-ing, licensing and construction compa-nies associated with each project, as well as the project’s name, status, location, cost, capacity, completion date, scope and project history. Users can access contact information for key project personnel.

For more information, visit www.constructionboxscore.com. Learn why Boxscore has been used for over 60 years by engineers, contractors, marketing professionals and business developers to identify construction projects around the world for lead generation, market re-search, trend analysis and planning.

HERMETIC-Pumpen [email protected] · www.hermetic-pumpen.com

Today’s Application: LNG-PROCESSING, HANDLING AND STORAGE

HERMETIC design features

100 % leakage free

Low life-cycle-costs

Low noise level

High reliability

Customized design – adapted to your process requirements

Customer‘s technical specifi cation

Capacity: 120 m³/h

Head: 1400 m

Pressure rating: PN 100

Motor power: 370 kW

H E R M E T I C A L L Y S E A L E D C E N T R I F U G A L P U M P S

HERMETIC-Pump Type CAMTV 52/6+6

Detailed and up-to-date information for active construction projects in the refining, gas processing, and petrochemical industries across the globe | ConstructionBoxscore.com

LEE NICHOLS is director of Gulf Publishing Company’s Data Division. He has five years of experience in the downstream industry and is responsible for market research and trends analysis for the global downstream construction sector. At present, he manages

all data content and sales for Hydrocarbon Processing Construction Boxscore Database, as well as all corporate and global site licenses to World Oil and Hydrocarbon Processing.


The Emerson logo is a trademark and service mark of Emerson Electric Co. ©2013 Fisher Controls International LLC. D352200X012 MZ8

Manually inputting the control signal feels pretty primitive. I need to get back in automatic mode for better efficiency.

YOU CAN DO THAT

You can automate your control signal using the Fisher® Control-Disk™ valve from Emerson. Process control loops containing butterfly valves

are often placed in manual mode due to poor control performance. This results in operators constantly monitoring and adjusting the control signal, significantly reducing efficiency. With a control range comparable to a segmented ball valve, the Control-Disk valve enables control closer to the target set point. This allows you to leave your control loop in automatic mode, regardless of process disturbances. With low maintenance requirements and sizes up to NPS 36, it’s time to put the Control-Disk valve in your loop. Visit www.Fisher.com/automatic to watch an animation video or download a brochure.

GE Works to redefine pump efficiencies.

In processing industries, GE’s SPS pumping

systems work to increase efficiency, reduce

environmental impact and provide versatile

pumping solutions. For example, our multi-stage

centrifugal SPS pumps are a widely-used, efficient

solution for circulation, transfer boosting and

disposal applications. Our SPS systems provide

enhanced pumping efficiencies by lowering noise

and vibration levels, decreasing construction

lead times and delivering improved reliability and

extended runtimes. With GE, you have a worldwide

support system ready to provide engineering,

field service or sales support.

Take a look at our results, and see how

GE can work for you.

Call +1 713 260 6731 or e-mail [email protected]

www.geoilandgas.com/sps



Viewpoint ROBERT L. DE MARIA

Maintenance Engineering Technical advisor, Dakota Gasification Co.

ROBERT L. DE MARIA is the maintenance engineering technical advisor at Dakota Gasification Co. in Beulah, North Dakota. He provides technical support on plant equipment reliability through inspections and monitoring, troubleshooting of equipment problems as well as failure analysis, corrective engineering and ensuring plant compliance on the mechanical integrity portion of the OSHA 1910 PSM Std. Mr. De Maria has held previous positions in plant reliability, project management, plant engineering, maintenance and utilities management. He has 43 years of experience in the energy, petrochemical and food-processing industries. His expertise is in rotating equipment reliability, which includes design, audit, selection, troubleshooting, modifications and monitoring. His present focus is on special projects (urea) and total plant reliability through asset data and information management. He received a BS degree in mechanical engineering from Stevens Institute of Technology and has authored numerous papers on rotating equipment reliability.

Consider integral-gear compressors in CO2 services

In 1998, the author began researching suitable technology for compressing 95 MMSCFD of “bone dry” (–100°F dew point) carbon dioxide (CO2) from 16.7 psia to 2,710 psia. The compressed CO2 is delivered via a 205-mi pipeline to opera-tions near Weyburn, Saskatchewan, Cana-da. The oil fields use the gas in enhanced oil recovery (EOR). As an added benefit to the environment, virtually all of the inject-ed CO2 is expected to remain permanently sequestered in the depleted oil fields long after these fields are abandoned.

Options. Three compression options were closely considered for this project:

Option 1: Motor + gear increaser + low-pressure (LP) compressor + medi-um-pressure (MP) compressor + high-pressure (HP) compressor

Option 2: Motor + gear increaser + LP compressor + MP compressor and motor + pump

Option 3: Motor + integral-gear com-pressor with 4 pinions (8 stages of com-pression) with a maximum pinion speed exceeding 26,000 rpm.

Implementing Option 3. In 1999, the Option 3 system was installed (FIG. 1). This unit has been operating for over 12 years; thus, significant data are available. The Dakota Gasification Co. confirmed that the integral-gear compressors in dry CO2 service could achieve several impor-tant benefits:

• Requiring the lowest capital cost• Reducing footprint of compressor• Highest total efficiency• Providing simplest design, with

high unit availability• Using a bull-gear-driven main lube

oil pump simplifies the lube oil system by eliminating oil-rundown tank

• Installing two 50% capacity 20,000 hp machines allows partial continuous production

• Using inlet guide vanes provides the highest compressor flow turndown.

Selection process. The manufactur-ers’ capability, experience and reputation were considered during the selection of the compressor train supplier. The op-erating company took steps to ensure project success in applying an untested design. Budgetary and personnel resourc-ing were allocated to achieve high avail-ability. A reliability design audit, lifecycle cost analysis and sub-supplier preference reviews were conducted. During this comprehensive audit process, the shaft seals received considerable attention, and the long-term reliability upgrades were the focus of many discussions. An agree-ment was reached on the final design.

Cost considerations finalized on using carbon-ring seals for all stages. However, the seal housing design would be capable of accepting dry-gas seals in the event that the carbon-ring seals were proven unsatisfactory. A factory-performance test with CO2 was done to confirm that aerodynamic performance was achieved. Of course, mechanical test-run data were captured and closely analyzed.

Key design elements. Other mechani-cal features under scrutiny included state-of-the-art “flexure pad” bearings, hydraulically fitted coupling hubs and diaphragm-spacer couplings at the driver connection. The motors are synchronous across-the-line starting design with liquid cooling. An original equipment manufac-turer-designed control system manages

FIG. 1. The integral-gear compressor in service for the Dakota Gasification Co.

Viewpoint

36�

speed and continuously monitors ma-chine condition. The control system is integrated into the plant’s digital control system (DCS), along with shaft vibration, thrust position and bearing-temperature monitoring—all are arranged for auto-mated alarm and shutdown.

Within the control system, the design needed to ensure that no liquid will form below the 1,100-psig compression sec-tions during startup and normal opera-

tion. Air-flooded interstage gas coolers are used. A glycol/water solution is circulated through the oil cooler and motor for cool-ing. Heat is rejected from the glycol/water solution through an air exchanger.

In 2006, an identical third integral-gear compressor was placed in service to support additional CO2 sales. All three machines were converted to synthetic lu-bricating oil; mineral-base lubricants had been used previously. The changeover was

prompted when it was discovered that mineral oil was being contaminated by certain sulfur constituents in the impure CO2 stream. Further evidences showed that the carbon-ring seals allowed gas to leak into the lubricant. However, the own-er’s reliability focus prompted rigorous in-service testing and comparison of differ-ent oils under varying load and operating conditions. With the synthetic oil, a mini-mum efficiency gain of 2% was realized. In addition, the new oil demonstrated higher resistance to degradation. The incremen-tal cost of the synthetic oil was paid back within a few months. Moreover, the syn-thetic oil has now been in service for over six years with no replacements necessary. The continued suitability of the synthetic oil is verified by periodic oil analysis.

Seal service life. The carbon-ring seal service life on HP stages 7 and 8 has met the manufacturer’s warranty. However, a measure of unscheduled downtime and recompression of higher-than-anticipated rates of CO2 has occurred. The events in-dicate a leak back to suction due to seal wear. The original seals did not reach the intended goal of 20 years between seal re-pairs. However, the carbon-ring seals for stages 1 through 6 have proven satisfac-tory. In 2010, the Dakota Gasification Co., in partnership with John Crane and MAN Turbo and Diesel, began the design and development of dry-gas seals for this ap-plication. These seals were designed, built and tested for the seventh and eighth stag-es. Installation is planned for June 2013.

Overall, these eight-stage compressors have met and exceeded expectations. The cost savings exceeded 10% as compared to the initial investment from the next-best option. The machines have achieved the lowest operating cost, along with 96% availability. With increased interest in CO2 sequestration projects, EOR and urea production, this technology presents merits for further consideration.

FIG. 2. Impellers for eight-stage compressor.

Problem SolvingSynthetic Lubricants

Reduce overheating, oxidation,excessive bearing and mechanical seal wear

Improve your pump reliability, extend MTBO*, reduce downtime andenergy consumption with Summit Syngear SH®-7000 Series and BarrierFluid Series. These synthetic lubricants are resistant to rust, oxidation,corrosion and improve wear protection. They have excellent lowtemperature fluidity and high temperature stability. Summit SyngearSH®-7000 Series and Barrier Fluid Series are compatible with mostprocess fluids being pumped and commonly used seal materials. ThesePAO based lubricants come in a wide range of viscosities.

Summit Barrier Fluids are NSF H1 Registered.

800.749.58239010 CR 2120Tyler, TX 75707 Made in USAwww.klsummit.com

Summit Industrial Products

*mean time between overhaul


[email protected] a Revolution Ahead!

Since 1968

VISIT US AT OTC - BOOTH #4929Select 54 at www.HydrocarbonProcessing.com/RS

| Special Report

MAINTENANCE AND RELIABILITY Hydrocarbon processing facilities are multi-billion-dollar industrial complexes designed

with service lives exceeding 30 years. To keep these plants operating as designed,

companies conduct regular maintenance and reliability programs to replace worn

equipment. Maintenance and reliability efforts are “insurance policies” to maintain

safe operations and yield high-quality products over the service lives of these facilities

and processing units. Such efforts reap numerous benefits.

Retrofit of valves for the tank farm operated by Turkish Petroleum Refineries Corp.’s

(Tüpraş’) refinery at Izmit, Turkey. Photo courtesy of Rotork Controls, Bath, UK.


Special Report Maintenance and Reliability J. BAILEY and S. BRADSHAW, ITT Goulds Pumps,

Seneca Falls, New York

Avoid hidden costs of suction-specific speed in pumping

More than 30 years ago, a landmark study observed a rela-tionship between suction-specific speed and the probability of pump failure that changed the way pumps are selected. This observation created the common perception that a lower suc-tion-specific speed value equals higher pump reliability, with a value below 11,000 established as a common benchmark for good reliability.

This perception has remained relatively unchanged, even though pump design and manufacturing methods have ad-vanced significantly. Pump purchasers continue to rely on a specification limit that derives from one study conducted by one company several years ago. If purchasers order the wrong design or more pump than is needed, then end users, suppliers and plant designers all experience economic consequences.

Here, suction-specific speed, which is one of the most com-monly mentioned (yet least understood) terms in oil and gas pumping, is broken down and analyzed.

The start of the specification limit. Referred to as “Nss” in the US or “S” in Europe, suction-specific speed originally helped pump designers predict and compare pump perfor-mances. Used since centrifugal pump theory was first devel-oped, Nss is not a speed at all, but rather a simple measure of a pump’s suction. It is based on the net positive suction head required (NPSHR) of a pump, and it should be calculated only at the pump’s maximum diameter and best efficiency point (BEP) flow:

(1) NSS =

RPM QNPSHR0.75

Where RPM = rotating speed of the pump, and Q = flow. Note: For double-suction pumps, divide Q by 2.

For a number of years, pump users sought to lower the el-evation of tanks to reduce piping costs and the expenses asso-ciated with higher-elevation equipment. This required pump manufacturers to strive for lower NPSHR values. Based on the formula in Eq. 1, as NPSHR decreases, Nss increases.

The relationship between higher Nss and lower reliabil-ity comes from the way designers traditionally achieved lower NPSHR—by increasing the diameter of the impeller eye. As the impeller eye grows larger, it results in increased suction recir-culation. Pump capacity is further reduced and the intensity of the circulation increases, causing a reversal of flow at the suction

pipe near the pump. If this effect is strong enough, it can cause cavitation (FIG. 1).

Numerous technical papers and articles reported the prob-lems caused by suction recirculation, but none quantified the problems until the aforementioned refinery—Amoco’s Texas City, Texas refinery—reported the results of a five-year study in 1982. The study involved nearly 500 centrifugal pumps, all of which were designed in the 1960s or earlier. The findings suggest-ed that, when pump Nss values exceeded 11,000, reliability halved.

From this, a commonly understood cap on Nss value was born. The 11,000 value was never promoted as an industry standard, but it became widely accepted by end users, engi-neering contractors and others as a specification to help ensure pump reliability.

Modern design and manufacturing techniques. Much work has been done in pump design and manufacturing to im-prove pump performance since the origin of the Nss limitation:

• More robust construction standards, as set forth in API 610, 8th edition, reduce the vibration effects of “off-best effi-ciency point (BEP)” operation

• Modern impeller design methods limit the need to in-crease the impeller eye diameter to achieve lower NPSHR

• The increased use of investment casting, ceramic core techniques and better mold washes have improved component surface finish, accuracy and repeatability.

d1

Recirculation

Impeller cross-section

FIG. 1. As the diameter of the impeller eye grows, it results in increased suction recirculation.

Maintenance and Reliability

40�

Most importantly, widespread use of computational fluid dynamics (CFD) allows pump designers to better define hy-draulic passages and improve vane profiles, instead of simply enlarging the impeller eye and hoping for the best outcome.

CFD is a sophisticated, computer-based modeling tool that can be used in the pump design process to simulate various de-signs, identify flow problems, develop solutions and evaluate operating strategies. As such, CFD is a cost-effective alterna-tive to physical modeling. With modern manufacturing tech-niques and materials, the companies that make pumps are bet-

ter able to produce the designs developed using CFD, enabling real-world tests on working pumps that are the ultimate test of any design technique.

A recent study sought to demonstrate the validity of one company’s CFD modeling approach. The results ultimately showed how a single, hard-and-fast limit on CFD can be coun-ter-productive. The study investigated how the suction perfor-mance of a pump was improved simply by changing the lead-ing-edge profile of the impeller vane while holding all other parameters constant. Four profiles were created—parabola, el-lipse, circular and blunt—using rapid prototyping techniques.

As shown in FIG. 2, CFD results matched physical testing fair-ly well; the experience was almost exactly as predicted by the CFD models. Variations only occurred at 120% best efficiency flow, which means that assumptions around the analysis tend to break down when there is extreme overload on the impeller.

It is important to note that CFD studies are not a “magic bullet”—they offer only a rough approximation of reality. The methodology should always be verified before CFD studies are used to select a pump. When choosing a vendor, several ques-tions must be asked about CFD studies:

• How many cells and what type of mesh (i.e., hexahedral or unstructured) were used?

• What turbulence model was used, and what conditions were applied to it?

• What were the inlet/outlet condition assumptions?• What was the residual error (root mean square and peak)

in the converged solution?• What distribution of Y+ values was obtained?• To what level of accuracy is the vendor willing to commit?

NPSH 3% head drop performance

0.1

0.15

0.2

0.25

0.3

0.35

0.7 0.8 0.9 1.0 1.1 1.2 1.3Q/QBEP

Cavit

ation

num

ber,

�3

Parabola CFDParabola test dataEllipse CFDEllipse test dataCircular CFDCircular test dataBlunt CFDBlunt test data

FIG. 2. CFD test results vs. physical test results.

8,000

9,000

10,000

11,000

12,000

13,000

14,000

15,000

16,000

0 1,000 2,000 3,000 4,000Pump Ns

Atta

inable

pum

p Nss

Attainable andacceptableperformance

Not attainablewith acceptableperformanceBetween

bearingdesigns

FIG. 3. Performance attainable using modern design techniques.

TABLE 1. Predicted impeller life based on impeller specifi cations

Impeller profi le

NPSH 3%, ft (m)

Nss/S, US units

Predicted impeller life* for Cast CA6NM (Bhn 262) hours

Blunt 36.8 (11.2) 10,386 15,208

Circular 33.4 (10.2) 11,170 16,631

Ellipse 30 (9.1) 12,104 22,971

Parabola 28.3 (8.6) 12,644 32,023

*Estimated service life of impeller pumping fresh water at 25°C with 23 ppm of dissolved gas content: BEP fl ow with � = 0.27 and inlet eye velocity = 30.4 m/s (99.7 ft/s).




• Has the CFD methodology been validated in lab tests?In general, the use of CFD allows pump manufacturers to

produce smarter solutions with higher reliability. Ultimately, these improvements benefit customers through lower cost and reduced risk related to pump station construction and operation.

Industry costs of relying on old specs. Despite these ad-vancements, persistent application of the Nss limit is slowing innovation. Purchasers continue to exclude more economic and reliable pumps from consideration because they are not compliant with the 11,000 Nss value.

For instance, the results of the aforementioned study showed that the parabolic leading-edge profile has the best cavitation performance. In fact, the expected impeller life is doubled between the parabola and the circular profile. For this application, a customer who specifies an Nss value of less than 11,000 would get a pump with poorer NPSHR performance and shorter life, with no improvement in vibration or mean time between failures (MTBF) compared to the design with a higher Nss number (see TABLE 1).

End users take on increased lifecycle costs from purchasing larger, slower and less efficient pumps. Plant designers suffer from reduced optimization because of bigger piping require-ments and higher tank elevations. The point is not that Nss is irrelevant, but that, with modern designs, manufacturers can produce reliable pumps with higher Nss numbers than were possible before.

A modern role for Nss in pump selection. When select-ing pumps, customers should work closely with consultants and pump vendors to understand the design options and the reliability performance of pumps for each application. If done properly, CFD simulations can provide good predictions of pump performance.

FIG. 3 shows the performance attainable and compliant with HI/API 610 vibration limits, using modern design tech-niques. While suction-specific speed is still a relevant measure for pump manufacturers to assess, setting a hard limit for this specification is no longer appropriate. Doing so places an un-natural constraint on the industry that limits pump users, sup-pliers and plant designers alike.

JOHN BAILEY is the oil and gas global product marketing manager for ITT Goulds Pumps, responsible for marketing its full range of products that serve customers in the oil and gas industry. Before joining ITT, Mr. Bailey spent 12 years in marketing and management roles for global companies serving industrial and technical markets. He completed GE Corporate’s Hands-On MBA program, and also holds an MBA degree from the University of Connecticut and

a degree in mechanical engineering from the University of Illinois at Champaign.

SIMON BRADSHAW is the director of API product development and technology for ITT Goulds Pumps. He has more than 20 years of engineering experience in the pump industry. Mr. Bradshaw also has supported the Hydraulic Institute in the development of pump standards and best-practice guides. He holds a BS degree in mechanical engineering from Heriot-Watt University, and is a registered chartered engineer in the UK and a member of the

Institute of Engineering Designers.


www.lindeus-engineering.com

Head Of ce: Five Sentry Parkway East, Suite 300, Blue Bell, PA 19422 USA 610-834-0300Texas: 3700 West Sam Houston Pkwy. South, Suite 425, Houston, TX 77042 USA 281-717-9090

[email protected]

For over 60 years, we’ve been there for our re ning and petrochemical customers. Linde Engineering North America Inc. offers single source responsibility for technology, engineering, procurement and construction.

• Selas Fluid re nery and petrochemical red heaters• Oxidation/incineration technologies• Engineered revamps and rebuilds• Hydrogen and synthesis gas plants• Air separation plants

A new name, a long history.

Selas Fluid Processing is now Linde Engineering North America Inc.

We’re local. We’re global.And we’re proud to be both.



Special Report Maintenance and Reliability D. RENARD, Elliott Engineered Solutions,

Jeannette, Pennsylvania

Rerating rotating equipment optimizes olefins plant performance

The operators of mega ethylene plants are under constant pressure to meet enhanced performance specifications and more stringent environmental regulations, while, at the same time, reducing energy costs and improving feedstock flexibil-ity. As ethylene plants have evolved in response to market de-mands (FIG. 1), technical advances in rotating equipment have boosted operating ranges and performance parameters for the centrifugal compressors and steam turbines—both are at the heart of ethylene plant operations. When considering the effect of changing market conditions on plant performance, particularly feedstock flexibility, producers must first evaluate changes to gas composition and process parameters in light of the plant’s existing turbomachinery.

Rerating installed compressors and steam turbines can be a cost-effective, time-saving solution for increasing throughput without investing in new equipment. Advances in flow path design, stage performance, aerodynamics, manufacturing technology and materials science make it possible to achieve new process parameters within the existing casings, with min-imal changes to foundations, piping and other connections. With careful planning, from review of process parameters through turnaround execution, equipment rerates can be ac-complished during a normal maintenance shutdown.

The presented case study will discuss typical cracked-gas (CG) and refrigeration services within an ethylene plant, highlighting technical advances that make it possible to achieve new process parameters with rerated equipment. The case study focuses on a CG equipment train installed in an ethylene plant in the late 1960s that is still in operation due to multiple rerates to meet new processing requirements.

Cracked-gas service. CG turbomachinery configurations vary from installation to installation. CG trains usually consist of large-volume capacity compressors driven by high-power steam turbines. Generally, CG trains have two, three or four compressor casings and sometimes a speed-increasing gear between the casings, as shown in FIG. 2. The most common ar-rangement is three casings. Three-compressor trains include a double-flow compressor for the low-pressure (LP) section. Three- and four-compressor arrangements provide more flex-ibility when considering process changes than a two-compres-sor train. In most cases, interstage pressures can be modified to accommodate the change in casing flow and head to opti-mize total performance.

Refrigeration service. Refrigeration compressors are sin-gle-casing configurations with multiple side streams or side loads and/or extraction streams, as shown in FIG. 3. Propyl-ene compressors are generally larger volume than the corre-sponding ethylene compressor. However, both applications require unique analysis when evaluating side-stream mixing performance to ensure that required interstage pressures and flows can be met. Refrigeration compressors have multiple nozzle connections that make the ability to reuse the casings a project objective.

1980

Plant

size

, tho

usan

d tpy

0

500

1,000

1,500

2,000

2,500

1990 2000Year

2010 2020

FIG. 1. Growth trend in ethylene plant capacity.

3 casings

ST driver MD

4 casings

ST driver LP 1stsection

LP 2ndsection

MD MD

2 casings

ST driver LP Gear MP/HP

KeyST–Steam turbineLP–Low-pressure compressorMP–Medium-pressure compressorHP–High-pressure compressor

HPLP doubleflow

FIG. 2. Typical CG train arrangements.



Advances in compressor technology. Centrifugal com-pressors, by design, cover a wide range of flow capacities. The development of impeller “families” has enabled a wider flow

range within a given compressor model. Since the 1960s, im-peller efficiencies have improved from 65% to nearly 90%, and the range of flow has doubled. This broadened flow range, coupled with the development of improved compressor stage head and efficiency, has dramatically expanded a given com-pressor’s operating range, as illustrated in FIG. 4.

Compressor design technology has advanced through the use of tools such as computational fluid dynamics (CFD),

finite-element analysis (FEA), solids modeling and rotor dy-namic analysis. These tools enable design engineers to model a three-dimensional view of the aerodynamic flow path and the

effects that design will have on overall performance.CFD analysis can be used to create impeller

stage ratings. Higher and lower flow stage ratings are derived from the tested components to form a family of stages. Within each family, impeller geom-etry is fixed. Blade heights are varied for higher and lower flows. Stage analysis results are continuously checked and verified against actual aerodynamic performance and field tests. This allows the design engineer to match impeller performance and sta-tionary diaphragm performance to achieve optimum

overall stage performance. By creating and extending impel-ler families, the application engineer can now select from sev-eral impeller designs to optimize stage-to-stage performance throughout the compressor aerodynamic flow path.

Refrigeration compressors usually have multiple side-streams that require accurate prediction methods. CFD analy-sis has enhanced the designer’s ability to accurately optimize the mixing of two flows while minimizing pressure drops for more reliable performance prediction. Propylene compres-sors represent a particular challenge when rerating an existing unit. In many cases, with multiple side streams, there may only be a single impeller in a specific section, for which the end user provides a pressure tolerance. The ability to select from various impeller families improves the likelihood that a solu-tion can be achieved.

Improving the performance of the entire flow path also requires consideration of all of the stationary components including the diaphragms, seals and casing volutes, as shown in FIG. 5. For example, diaphragms are milled and bolted to eliminate rough surface finishes inherent with older cast tech-nology. Interstage sealing is accomplished with abradable and deflection-tolerant interstage seals to maintain efficiencies for longer periods. These advancements in stage performance make it possible to reuse existing compressor casings with minimum impact on the equipment train driver. Improve-

Rerating installed compressors and steam

turbines can be a cost-effective, time-

saving solution for increasing throughput

without investing in new equipment.

0.00.6

0.65

0.7

0.25

0.8

0.85

0.9

0.05 0.1Flow coefficient

Impeller efficiencies(approximate)

Mid-1990s ––> Today1980sMid-1960s and 1970s1950s and 1960s

Polyt

ropic

efficie

ncy,

%

0.15 0.2

FIG. 4. Evolution of impeller performance. FIG. 5. Typical compressor component upgrades.

FIG. 3. Refrigeration compressor with multiple side streams.



ments in manufacturing technology have also contributed to improved compressor design performance. For example, five-axis milling techniques allow high-performance impel-ler blades to be used for higher flow and/or head to expand a compressor’s operating range.

Advances in steam turbines. Significant plant capacity increases that result in higher compressor flow—and, there-fore, power—cannot be achieved without a correspondingly significant increase in steam flow. Steam turbine drivers must use the existing steam operating conditions to match the re-rated compressor train speed. Casing size limitations provide unique challenges to reconfiguring the existing steam flow path to achieve the desired power and speed.

CG and propylene drivers are usually high-power units that may require significant changes when the compressor power has increased. Even with improvements in stage effi-ciencies over the years, an increase in power usually requires an expanded steam flow area inside the turbine, which may or may not be possible within the physical limits of the existing casing. Additional modifications to the existing aerodynamic flow path, such as removing stages, are usually required to in-crease the steam flow area.

CFD analysis and other analytical tools have advanced performance and reliability in turbine component design, specifically rotating blades and stationary diaphragms. CFD analysis on high-pressure (HP) and LP staging is used to op-

timize both the rotating and stationary turbine components. Additional improvements include replacement of labyrinth seals with brush-type seals and tip seals (FIG. 6).

Case study. This case study references an ethylene plant that was originally built in the late 1960s with a two-body CG compressor train, as shown in FIG. 7. Over the years, the LP compressor was rerated three times prior to this project, and the HP compressor and steam turbine driver were rerated

FIG. 6. Typical steam turbine component upgrades

We Process Your Energy

Because Oil & Gas projects require highly reliable and cost effective solutions we timely deliver since over 35 years fully guaranteed modular process-equipment packages tailored to meet specific customer’s requirements; for all kind of Oil & Gas processing applications.

Because each project is unique and highest quality design is paramount to success we offer a unique portfolio of proprietary gas treatment technologies licensed by Prosernat. Backed by our strong investment and partnerships in R&D, with IFP Energies nouvelles and Total, we are proud to bring to the market the most advanced technologies in the following areas: Natural Gas Sweetening, Natural Gas Dehydration, Sulfur Recovery, Hydrocarbon Dewpointing, Separation and CO2 Capture from Flue Gases.

TECHNOLOGIES AND EQUIPMENT FOR OIL & GAS TREATMENT YOUR EFFICIENCY, OUR FLEXIBILTY

p jwe timely deliver sincequipment packages taall kind of Oil &Gas pro

Because each project issuccess we offer a uniqlicensed by Prosernat.in R&D, with IFP Energthe market the most adGas Sweetening, NatuDewpointing, Separatio




twice. These earlier rerates increased plant operating flow and power by 55% over the original installation specifications.

Recently, the end user decided to change the plant’s feed-stock and to expand the plant capacity by an additional 58%. The owner asked the compressor’s original equipment manu-facturer (OEM) to conduct a feasibility analysis. The OEM worked closely with the end user, the engineering contractor and the ethylene process licensor to evaluate process condi-tions and develop a solution.

Early in the analysis, it became clear that such a large flow increase would require installing a booster compressor to re-duce the volume flow into the existing units. Reducing the vol-ume flow to the LP and HP compressors allowed these units to be rerated yet again for a fourth and third time, respectively.

To meet the required increase in flow and to accommodate the feedstock revision, the LP and HP compressors required all new rotors and stationary components, but with a reduced number of stages in each unit.

Advanced high-performance impeller technology was used to achieve the expanded flow requirements while limiting the overall train power to an increase of approximately 35%. This allowed the steam turbine driver to be rerated for the third time. The steam turbine also required a new rotor and new dia-phragms, with a reduced number of stages.

The end user was able to achieve its operating objectives for capacity increase and feedstock flexibility while minimizing site work and reusing the existing casings, with minimal investment in new turbomachinery hardware. From the original plant in-stallation in the late 1960s, the total train power has increased by more than twice the original design. Application of the latest in compressor and steam turbine technology has enabled the existing compressors and steam turbine to remain in operation for nearly 50 years.

DAN RENARD, commercial operations manager retired, spent 28 years at Elliott Co. in various management positions in the compressor and turbine application engineering and marketing departments. He holds a BS degree in mechanical engineering from the Pennsylvania State University. Until his retirement in 2012, he was the marketing manager for the Engineered Solutions Group located at Elliott’s US headquarters in

Jeannette, Pennsylvania, leading a team responsible for the rerating of turbomachinery equipment worldwide.

Finally, a Safety Lifecycle Management Solution...

Powered By

aesolns.com

aeShield™ Safety Lifecycle Management System is a comprehensive platform for executing a sustainable risk management program through automation of the safety lifecycle process. The system provides a complete solution by maintaining relationships among the risk reduction targets, design verification calculations, inspection and test plans for integrity management, and actual historical data. aeShield tracks and analyzes PSI, providing alerts and reports on process safety health in real time. aeShield facilitates work flow and compliance with ISA84.00.01/IEC 61511 and the related requirements of OSHA 1910.119.

g

Shielstemecutiogramecycleompation

rgetspecti

anageShiel

erts areal tiw an

C 615OSH

aeSSysexeprolifea crelatat rinnsmaaea Sala ein rflowIECof O

Shie

ldTM

Shie

lddTTMM

DESIGN, IMPLEMENT, & CONSTRUCT

WITH the experts to implement it.

ST driver Booster LP

Before

After

MP/HP

ST driver LP Gear

Gear

HP

FIG. 7. Rerate case study: Before and After.


Hazardous events pose serious threats to operat-

ing companies by negatively impacting reputation and

commercial performance as well as endangering people

or the environment. Appropriate test plan strategies

provide an important validation that safety systems

and protective systems are functioning correctly. Test

documentation can be time consuming to maintain, in-

effi cient for technicians to work with, and diffi cult to

connect test results back to the original drivers.

aeShield™ provides enhanced connectivity between

risk assessment and testing strategies to allow operating

facilities to make the best use of limited resources. Im-

proved connectivity between data allows for better deci-

sion making on test strategies and improved tracking of

test performance.

SIF Testing aeShield optimizes strategies for Safety Instru-

mented Function (SIF) test plans including online de-

vice level testing and offl ine full SIF level testing. Test

plan template management within aeShield reduces te-

dious eff orts of updating testing plans manually and en-

sures current test plans include accurate SIF function-

ality. Th e Look Ahead feature illustrates which tests

need to be completed in upcoming months to improve

decisions about when it is most eff ective to conduct

those tests. By emphasizing drivers for testing and doc-

umenting required SIF functionality, aeShield provides

maintenance groups a cohesive platform for managing

test plan strategies.

IPL Testing aeShield increases visibility of how assumptions

made during the HAZOP/LOPA process have sig-

nifi cant eff ects on testing requirements. Independent

Protection Layers (IPLs) must be validated to ensure

accuracy of selected SIL targets. aeShield links the data

from the HAZOP/LOPA to the IPL instrumentation

model, enabling automatic notifi cation when additional

IPL tests are needed.

Test Data Collection aeShield off ers multiple ways to collect and main-

tain electronic data. aeShield queries existing asset or

calibration systems already being used to store test re-

sults and links the test results from the external system

to design information. By linking test results and de-

sign information back to the initial requirement data,

aeShield provides the mechanism to compare real-

world test data to assumptions made in front end load-

ing and design activities.

Th e aeShield import utility enables users to batch

import data from a fl at fi le, manually enter data directly

into aeShield, or use both mechanisms. Th is fl exibility

enables less time spent entering data and improves data

availability.

By establishing connectivity between risk assess-

ment, instrumentation design models, and real world

data collection, aeShield improves decision making

for testing strategies and provides tracking of test per-

formances. Organizations can select from a variety of

implementation options and services that best fi t their

needs. As an industry leader in process safety and SIS

design and implementation, aeSolutions is prepared to

assist you at every step of the Process Safety Lifecycle,

helping achieve compliance and sustainability.

Maintenance & Reliability: Connectivity of Data for IPL & SIF Testing. Are you making the best use of your data?

MIKE SCOTT / Mike.Scot [email protected]

A D V E R T O R I A L

Use your QR code reader to learn more about aeShield – Safety Lifecycle Management or visit www.aesolns.com.

Test Plan Look AheadTODAY’S DATEApril 3, 2013 Advance Month >

April 2013 May 2013S M T W T F S

1 2 3 4 5 6

7 8 9 10 11 12 13

14 15 16 17 18 19 20

21 22 23 24 25 26 27

28 29 30

S M T W T F S

1 2 3 4

5 6 7 8 9 10 11

12 13 14 15 16 17 18

19 20 21 22 23 24 25

26 27 28 29 30 31

1. [AE-101-100]-[PT-100, PT-200, PR-300]-[3]2. [AE-101-117]-[TT-100, TT-200]-[2]3. [AE-101-096]-[FF-100]-[1]4. [AE-101052]-[LT-100]

IL RATING 1 2 3

SELECTED DATEApril 26, 2013

FIG. 1. Test plan look ahead reporting.

FIG. 2. Easily identify overall status of facility test plans.

Execution Status

SIF-LEVEL TEST PLANS DEVICE-LEVEL TEST PLANS

Complete

Incomplete

Complete

Incomplete

OVERALL STATUS

Early On-Time Late At Risk

25% 50% 75%

Complete Incomplete


KOBELCOOil-Free Screw Compressors –

A BrilliantSolution for Dirty GasesFlare and Vapor Recovery ServiceKobelco oil-free screw compressors are the clear solution for heavy, complex, corrosive, unpredictable gases. They don’t need oil in the compressor chamber, so there’s no risk of contamination or breakdown in viscosity. Best of all, they compress any gas and deliver years of continuous, uninterrupted operation.

High-Capacity Oil-Free Screw - The world’s most sophisticated Oil-Freescrew compressor, with the largest capacity – up to 65,000 CFM (110,000m3/hr). Ideal for refinery flare gas recovery and petrochemical polymer forming gas.

Advanced Oil-Free Screw - Compact design conserves valuable space onoffshore platforms and FPSO’s and in refineries. Handles VRU, Flash Gas, LP and MP flare services, accommodating fluctuating gas compositions with heavy hydrocarbons, H2S and water.

Kobelco Compressors America, Inc.Houston, [email protected]

Kobe Steel, Ltd.Tokyo +81-3-5739-6771Munich +49-89-242-1842

www.kobelcocompressors.com

Ask Kobelco! The Best Solution for Any Gas Compression.



Special Report Maintenance and Reliability T. MAYNE and M. ELLUL, Qenos Olefins Australia,

Melbourne, Australia; and D. PHILLIPS, Emerson Industrial

Automation, Hanover, Maryland

Maximize steam unit performance with precise torque monitoring

All turbomachinery is subject to degradation that, over time, will affect the system’s efficiency and operational perfor-mance. Precise monitoring of turbomachinery performance with continuous torque-monitoring systems can be used to identify gradual efficiency loss, allowing for the development of a more focused maintenance scope to return the system to its optimum operation and efficiency.

Torque monitoring based on heat balance, energy balance and other methods utilizes numerous parameters such as pressure, temperature, flowrate, gas composition, etc., which require precise instrumentation to measure with low uncer-tainty.1 However, phase displacement technology can be used to accurately measure torque directly at the coupling to within 1% of full-scale torque, a combination of all electrical and me-chanical sources of error. This accuracy provides the lowest amount of uncertainty when computing efficiency, compared to alternative methods.

A torque-monitoring system was recently installed on a cracked-gas compressor (CGC) train at Qenos Olefins in Australia to determine the causes of a power limitation. The torque-meter coupling utilizes phase displacement technology for long-term reliability, eliminating the need for recalibration.

Torque meter installation. The meter consists of two rings with pickup teeth installed on a torsionally soft spacer and inter-meshed at a central location. Two monopole sensors 180° apart are mounted on the coupling guard. As the coupling rotates, the ferromagnetic teeth create an AC voltage waveform in the sen-sor coil, which is digitally processed using known calibration parameters. Due to the intermeshed pickup teeth, the system is referred to as a single-channel phase displacement system, producing two independent torque measurements (FIG. 1). The system will output torque, power, speed and temperature, which can be easily integrated with any DCS system (FIG. 2).

At the olefins plant, the operating cycle of the steam-driven, CGC train is 7–8 years. During this cycle, the plant reaches production limitations because this compressor train encoun-ters a power limit. To determine the cause of the power limit as “turbine fouling” or “compressor fouling”—or a combination of both—was not confidently possible with the instrumenta-tion installed.

One option was investigated to add more power by upgrad-ing the turbine power rating from 7.5 MW to 9 MW. This re-

quired a capital investment of $2 million. The plant elected to defer this investment and, instead, a torque meter was installed during the major eight-year shutdown.

The installation involved replacing the existing coupling spacer and flexible halves with the “drop-in” torque meter’s integral flexible elements. The torque meter assembly was dynamically balanced to API standards, so it was not neces-sary for the user to return any coupling components for the retrofit. The coupling guard was modified so that the two variable-reluctance sensors could be installed, completing the mechanical installation (FIGS. 3–5).

Volta

ge

Time

Sensor 1Sensor 2

4

3

2

1

0

-1

-2

-3

-4

FIG. 1. The torque-meter coupling produces two independent torque signals.

Value

(N-m

, kW

, rpm

)

Time

TorquePowerSpeedTemp

Tem

pera

ture

, °C

60,000

50,000

40,000

30,000

2,0000

10,000

0

60

50

40

30

20

10

016:37:11 16:40:42 16:44:12 16:47:44 16:51:12 16:54:43 16:58:13 17:01:44

FIG. 2. Typical output from the torque-meter coupling.



Results. On restarting the plant and having completed a num-ber of compressor efficiency improvements, the torque meter clearly showed that the 7.5-MW turbine did not require an uprate and that the major power losses were coming from the

CGC. The torque meter also allowed online tuning of the seal gas system of the compressor to establish the lowest power draw from the recycles that the seal system introduces. An ad-ditional 200 KW of power was reduced from the turbine load, with the manual adjustments made on the seal gas system.

The torque meter is now being used to monitor turbine steam-fouling issues and process-related compressor fouling

so that corrective online washing can be activated as soon as issues arise.

The historical data collected from the torque meter will also provide a baseline of mechanical loading through the

drivetrain of the CGC over time. This data will be used to determine if increases in the maximum continuous operating speed rating of the compres-sor and the turbine can be accomplished at minimal costs. This would achieve increases in the operating envelope of the compressor.

Furthermore, the value of the torque meter justi-fied the installation of a second system for the ole-fins plant’s second steam-cracking plant turbine/compressor train in October 2012.

LITERATURE CITED 1 Kurz, R., K. Brun and D. Legrand, “Field performance testing of gas turbine-driven

centrifugal compressors,” Proceedings of the 28th Turbomachinery Symposium, Turbomachinery Laboratory, Texas A&M University, College Station, Texas, pp. 216–220, 1999.

DANIEL PHILLIPS is the field service engineering manager for Emerson Industrial Automation’s Kop-Flex brand of couplings in Baltimore, Maryland. He assists users with installation, commissioning and troubleshooting of power transmission products. Mr. Phillips has extensive experience with applying torque-monitoring solutions to increase the reliability and efficiency of equipment in the metals and oil and gas industries. He

has a BS degree in mechanical engineering from the University of Maryland, Baltimore County, and he has 10 years of experience in the mechanical engineering field.

MARK ELLUL has worked in Qenos Olefins Australia’s olefins refinery for 30 years as an instrument and electrical specialist. He has coordinated field maintenance activities and worked in the process and control applications group. Mr. Ellul has also been assigned to major rotating machinery instrument upgrade projects.

TREVOR MAYNE is the lead machinery engineer for Qenos Olefins Australia’s olefins refinery. He has worked with rotating equipment in the olefins refinery and in the plastics and synthetic rubber plants over the last 20 years. Mr. Mayne has held positions in reliability and in field maintenance, both at Qenos Olefins Australia’s Altona plant and in Saudi Arabia with ExxonMobil.

FIG. 5. Existing coupling arrangement (top) and retrofitted torque-meter coupling (bottom).

Precise monitoring of turbomachinery

performance with continuous torque-

monitoring systems can be used

to identify gradual efficiency loss,

allowing for the development of a

more focused maintenance scope.

FIG. 3. Completed mechanical installation at Qenos Olefins.

FIG. 4. Torque-meter coupling retrofit at Qenos Olefins plant.

Challenging times and changing technology call for forward-thinking solutions. That’s why we’ve taken our proven Mobil SHC™ synthetic lubricants — the standard-setting oils and greases for more than 40 years — into the future, unleashing the next generation of productivity with three new advances. Each delivering overall balanced performance with substantial energy-efficiency benefits.

Mobil SHC™ 600 Series — The enhanced formulation improves viscosity and low-temperature properties as it delivers outstanding performance across a wide range of circulating and gear applications. Mobil SHC™ Gear — This supreme gear oil was reengineered to deliver optimum equipment oil life in gearboxes, even under extreme conditions, with significant reduction in energy consumption.Mobil SHC™ Gear OH Series — Our customized formulation for the Off-Highway sector features dependable technology with excellent low-temperature performance.

Discover the advanced technology of Mobil SHC. To see how four decades of synthetic innovation just jumped a generation ahead, visit mobilindustrial.com.

The world of synthetic lubricants just took three giant leaps forward.

The energy efficiency design is a trademark of Exxon Mobil Corporation. Energy efficiency relates solely to the fluid performance when compared with conventional reference oils of the same viscosity grade in gear applications. The technology used allows up to 3.6% efficiency compared with the reference when tested in circulating and gear applications under controlled conditions. Efficiency improvements will vary based on operating conditions and applications.

© 2013 Exxon Mobil Corporation.All trademarks used herein are trademarks or registered trademarks of Exxon Mobil Corporation or one of its subsidiaries.


Select 55 at www.HydrocarbonProcessing.com/RSRAISING PERFORMANCE. TOGETHER™

Integrally geared compressors may be new to some, but we’ve manufactured over 12,000

since 1955. They’ve proven their reliability, fi rst for industrial gas applications, now for natural gas

compression and process applications, including small-scale LNG. The simplicity of design means

fewer working parts, more uptime and lower all-around costs. Find out more at www.c-a-m.com/cs.

Cameron’s Process Gas Advantages:

F L O W E Q U I P M E N T L E A D E R S H I P

PROCESS GAS APPLICATIONS

I N D U S T R I A L K N O W-H O WGEARED TOWARD

AD00523CC


Special Report Maintenance and Reliability N. GHAISAS, PE, Fluor Canada Ltd., Calgary, Alberta, Canada

Investigate power limitations in a large steam turbine

An investigation was carried out to rectify power limitations for a steam-turbine driver of a cracked-gas compressor (CGC) in a world-class ethylene plant.

Description. The turbine is a nine-stage, impulse-type, multi-valve, straight-condensing turbine. Five years after the initial start up, the first-stage nozzles of the turbine were replaced with a new set of nozzles supplied by the original equipment manufacturer (OEM). The replacement was done to yield higher power output now required for olefins produc-tion. The new nameplate-rated power of the turbine is 54,000 hp at 3,927 rpm. The revamp excluded any modifications to the compressor drive and couplings due to adequate design margins that could accommodate higher flowrate, pressure and torque.

Steam is supplied to this turbine from a high-pressure header at 700 psig at 900°F through a 24-in. nozzle. An in-verted, oil-operated trip and throttle valve is mounted just up-stream of the turbine. The top half of the steam chest contains five venturi-type, single-seated governing valves and valve seat nozzles. Each valve has a nominal diameter of 4 in. The valves are sequentially lifted by a bar mechanism. The I/P (current to pressure) converter of the actuator receives a 0 mA–20 mA signal from the turbine’s electronic governor in response to speed changes.

For nearly three years after the revamp, the turbine was in continuous operation and exhibited normal performance. However, first signs of deviation from steady-state conditions became apparent when the turbine was no longer able to de-liver power required by the CGC train. Turbine speed could not be maintained and had gradually reduced by as much as 5% from the rated speed of 3,927 rpm. Consequently, the set-point for first-stage suction pressure of the compressor had to be changed to increase the pressure. The servo-motor piston in the turbine governing system was at a fully retracted position, indicating that all five of the bar-operated valves were lifted. The situation became critical because the CGC’s capacity re-duction directly impacted ethylene production, and, ultimate-ly, negatively impacted earnings for this facility.

Source of power limits. Two possible reasons for the tur-bine’s power limitation were initially investigated and included:

• Deposit buildup on the rotor from steam impurities• Detached row(s) of blades.An analysis of the steam quality had been done on a month-

ly frequency. Test results did not indicate that impurities were present in live steam. Turbine vibrations were low (15 mi-

crons p-p on both radial bearings, and there was no cognizable changes in the thrust position of the rotor. Thus, the listed root causes were ruled out. Instead, as suggested by the author, a nozzle-bowl pressure survey was conducted to identify any ab-normalities inside the steam path.

With the governor valve lift indicator showing a fully open lift bar, steam-inlet pressure, nozzle-bowl pressures and first-stage pressure (also known as nozzle-ring pressure) were re-corded. TABLE 1 lists details from the nozzle-bowl survey. Refer-ence was also taken to the “steam-throttle flow vs. first-stage pressure” and “steam-throttle flow vs. valve lift” charts supplied by the turbine manufacturer.

Interpretation of the measured data revealed that:• Pressures in nozzle bowls 1, 2 and 4 were nearly equal to

the steam inlet pressure, but pressures in bowls 3 and 5 were found to be equal to first-stage pressure, i.e., the downstream pressure. FIG. 1 shows the arrangement of governing valves.

TABLE 1. Measured parameters

Measurement Value

Steam inlet pressure, barg 46.1

First-stage pressure, barg 39.8

Nozzle bowl No. 1 pressure, barg 46

Nozzle bowl No. 2 pressure, barg 45.8


Nozzle bowl No. 4 pressure, barg 46


Venturi nozzleSteam chest

Governor valve

42

13

5Steam chest internal

Inlet steam

Inlet nozzleValve seat

Valvelifter bar

Threaded endDouble nut

FIG. 1. Side view of the governing valves for the turbine.



• Since the pressure in nozzle bowls 1, 2 and 4 were in the proximity of steam-inlet pressure, the possibility of restricted steam flow due to a clogged strainer in the trip and throttle valve was eliminated. Similarly, since the governor lift rod had traversed to full lift position, binding or interference in gov-ernor linkage was not considered as a contributing factor to the problem.

• These results inferred that the power limitation was attributable to valves 3 and 5, in that these valves had likely detached from the lifting bar and blocked the venturi nozzles underneath.

Following the pressure survey, the turbine was shut down to open the steam chest top cover for internal inspection (FIG. 2). Upon disassembly, the stems of governing valves 3 and 5 were found to have sheared along the valve stem threads. The broken valves were sitting on the venturi nozzles, thus

blocking the nozzle passages. This explained the reduced steam flow to the turbine and the resulting power limitation.

The venturi nozzle for valves 3, 4 and 5 were found to be loose in the steam chest. The nozzles were originally set in bored openings in the steam chest with a tap fit and secured by one set screw.

During the shutdown, a boroscopic inspection of the rotor was conducted to confirm that there was no deposit buildup on the blades.

Following inspection, all five governing valves were replaced with a set of spare valves. Loose noz-zles were secured into place and locked with three set screws to prevent them from coming off. Tack welding of nozzles was not carried out due to the possibility of thermal distortion of the steam chest. After repairs and calibration checks of the govern-ing system, the turbine was warmed up. This was followed by a over-speed-trip test and then the tur-bine was coupled to the compressor string. FIG. 3 is the calibration curve for the actuator.

Further evaluations were carried out to investigate why the two governing valves had broken. These valves are the closest to the steam-inlet opening in the steam chest. Thus, they are first valves to take full steam pressure. At the time of revamp, only the first-stage nozzle ring was replaced, but, apparently, no mechanical design check was done by the manufacturer on the upstream steam path. It was determined from calculations that the steam velocity at the turbine inlet nozzle, corresponding to post-revamp conditions, was in the proximity of 150 ft/sec.

Although it is not an absolute value, good engineering guides recommend flow velocity through nozzles to be less than 150 ft/sec for reasonable pressure drop and flow dis-tribution. Continuous, prolonged operation at high-steam velocity and the associated energy of trapped steam can create turbulence in the steam chest. Such turbulence must have loosened the securing nuts on governing valves 3 and 5. The steam velocity worked its way through the weakest link, which is the valve threads, and then sheared the threads.

The findings and evaluation were communicated to the turbine manufacturer. The manufacturer concurred with the analysis and asked that the broken valves be returned for metal-lurgical investigation. In the end, the new set of replacement valves was nitrided at the manufacturer’s facility to create a case-hardened surface on ASTM A410 valve material. In the next available opportunity, these valves were installed in the steam chest. Since this changeout, the turbine has operated normally and is able to meet the process conditions demand.

NEETIN GHAISAS is a Fluor Fellow for rotating equipment and a director, design engineering in Fluor’s Calgary, Alberta, Canada, office. He holds an MS degree in mechanical engineering and is a registered practicing Professional Engineer in the province of Alberta, Canada. Mr. Ghaisas possesses over 31 years of professional experience, especially in the specification, selection, application and troubleshooting of rotating equipment. Mr.

Ghaisas is a subject matter expert for Fluor Corp. on compressors, steam turbines, reliability-centered maintenance and root-cause-failure analysis. Further, he has a number of years of experience in machinery vibration diagnosis including transient and steady-state analysis. In Fluor’s Calgary office, he serves as a group leader for rotating equipment engineers. Mr. Ghaisas is a member of the task force member for several API standards and is also a member of Machinery Function Team for Process Industry Practices.

Operation of the cracked-gas compressor

is a major factor governing the profitability

of any ethylene facility. Limitations on the

turbine driver for this compressor can be

the difference between profitability

or doom for any olefins site.

FIG. 2. Steam chest under investigation.

Signal, %

Lift,

%

0 10 40 60 80 100

20

60

100

FIG. 3. Calibration curve for the actuator.

Among these innovative products is a line of internals

media and to provide a collection area for the process

desulfurizers

gas sweeteners

www.johnsonscreens.com

JAPAN

CONTACT US!

Johnson Screens designs and manufactures a wide range of vessel internals for media retention.


Solutions That FitSM

Total Safety has the people, programs and processes to help mitigate risks during critical operations. Our skilled safety professionals have expertise in designing custom solutions that fit your specific needs, ensure safe operations, and protect your people and assets. Look to our dependable safety and compliance solutions to create the ideal working environment for your workers worldwide.

Watch our new Centralized Confined Space Monitoring System video on our website at TotalSafety.com.

Compliance Systems

Safety Personnel

Professional Services888.448.6825 | TotalSafety.com



Special Report Maintenance and Reliability S. SHAW, AESSEAL plc, Rotherham, South Yorkshire,

United Kingdom

Seal safety may require going beyond typical standards

In 1984, a single, large-scale emission of methyl isocyanate (MIC) led to the loss of thousands of lives at Bhopal, India. The event became one of the most consequential process safe-ty failures for the manufacturing industry. Follow-up releases and articles have documented how machinery failures played a role in the Bhopal disaster.1 Incremental changes made to the original process design involving pumps and seals ultimately contributed to this catastrophic event.

A comprehensive examination of current seal standards indicates that possible inconsistencies and weaknesses are present in these guides. Such deficiencies were confirmed in leakage testing conducted by a leading seal manufacturer in mid-2012. There is consensus among respected industry professionals that users must reach beyond typical standards in applying safer mechanical seals when hazardous and toxic chemicals are involved.

Pumps, seals and what happened in Bhopal. In the first major change from original plant design, engineers and manag-ers at the Bhopal facility adopted an alternative fluid-transfer method. The operators pressurized the MIC storage tank and reversed the flow directly into the derivatives unit, thus avoid-ing the failure-prone MIC-transfer pumps. According to im-portant cross-references released to the public record in early 2012, this alternative practice was adopted because it “mini-mized the potential for transfer-pump seal failures and expos-ing employees to the lethal process.”1

The second major change involved a “circulation pump” processing the MIC through a fluorocarbon-based refrigera-tion system. Refrigeration was used to keep the MIC tempera-ture near 0°C, thus greatly reducing the risk of a thermal run-away reaction. On January 9, 1982, there was a severe incident in which a mechanical seal face had shattered. A significant MIC release occurred, and 25 employees were sent to hospitals with serious injuries. The refrigeration system was no longer used after January 12, 1982.

The reasons for making process changes were identical. Both operational changes were implemented to avoid future pump and seal failures that could expose employees to lethal chemicals.

In May 1984, an independent audit team arrived at the Bho-pal MIC facility. The team made several recommendations, one of which was to install dual seals on the centrifugal pumps. Dual seals (FIG. 1) and appropriate seal-support systems or flush-plan arrangements can effectively control fugitive emis-

sions. In this case, the dual seals and support systems were not installed before the tragic release event on December 2, 1984.

Prevailing industry standards. In all hazardous services, seal-ing safety and equipment reliability interact. Present practices for sealing volatile compounds are linked to American Petroleum Institute (API) 682/ISO21409, which is the international stan-dard most often used. The standard specifies the requirements and offers best-practice recommendations for sealing centrifu-gal and rotary-positive displacement pumps. The hydrocarbon processing industry considers this standard of critical impor-tance in hazardous, flammable and/or toxic services.

It is widely understood that shaft sealing systems conform-ing to API 682/ISO21409 must meet four stated reliability ob-jectives, which are:

1. All seals should operate continuously for 25,000 hours without the need for replacement.

2. Containment seals (wet- and dry-mechanical seals) should operate for at least 25,000 hours without the need for replacement at any containment seal chamber pressure equal to or less than the seal leakage pressure switch setting (not to ex-ceed a gauge pressure of 0.07 MPa/0.7 bar/10 psi), and for at least 8 hours, at the seal chamber conditions.

3. All seals should operate for 25,000 hours without the need for replacement while either complying with local emis-

FIG. 1. With a dual mechanical seal, the space between the sleeve and inside diameter of the two sets of seal faces is filled with a pressurized barrier fluid.



sions regulations, or exhibiting a maximum screening value of 1,000 ml/m3 (1,000 ppm by volume), as measured by the US EPA Method 21, whichever is more stringent.

4. The minimum performance requirements and permit-ted leakage detailed in Section 10.3.1.4.1 of the standard speci-fies an average liquid leakage rate of less than 5.6 g/hr per set of seal faces.

The fourth point is very significant. Because 5.6 g/hr is the maximum acceptable leakage rate given in this standard, and seal users expect compliance. Many users base their safety risk analyses on this information. However, mechanical seals can enter the supply chain and vastly exceed the presumed maxi-mum leakage rate of 5.6 g/hr.

Essential testing protocol and results. Testing for liquid leakage by actually using liquids is not always convenient. Test-ing with clean air is feasible as long as the equivalencies are estab-lished by calculation, calibrated tests and measurements. With this goal in mind, API 682/ISO 21409 allows seal manufactur-ers to follow an air-testing protocol. The protocol is described in Section 10.3.4.2 of the standard and is summarized in TABLE 1.

However, while meeting the air-test criteria, a mechanical seal can utterly miss the liquid leakage limit given in objective 4, from Section 10.3.1.4.1 of the standard. That objective, as stated above, is to limit allowable liquid leakage from new mechanical seals to

5.6 g/hr. This potentially far-reaching safety issue prompted a thorough examination by calculations. The accuracy of the cal-culations was then validated by air-testing in accordance with the API protocol. Air tests were followed by water tests.

TABLE 2 summarizes the air-test results. Using the criteria from TABLE 1, air leakage through a 0.008-in. orifice in a pressur-ized 26.5-l loop would pass the test. But the tests also confirmed that seals in a pressurized lower-volume loop (1.5 l) fitted with the same orifice would not meet the stipulations of TABLE 1. In a 1.5-l loop, the pressure loss was 22 psi in 5 minutes. Result: The seal failed the test. On the positive side, the test descrip-tions and results of TABLE 2 indicated that a well-designed and properly manufactured mechanical seal is virtually airtight. It will experience no pressure decay in either the small (1.5 l) or large (26.5 l) test loop.

Testing for liquid leakage. For testing with water in the seal loops, the loops were fitted with a 0.2-mm (0.008-in.) orifice. Water was added to the loop and connected to a pressurized volume of nitrogen. Three different nitrogen pressure settings were used for the water-test sequence, as listed in TABLE 3. All water tests showed that the liquid-leakage rates exceeded the maximum permissible limit of 5.6 g/hr:

• Filled with water and pressurized to a gauge pressure of 0.2 MPa (2 bar, or 29 psig), a mechanical seal actually leaked 1,500 g/hr—270 times the allowable rate.

• When pressurized to 7 barg (102 psi), the leakage rate jumped to 50 g/min or 3,000 g/hr (6.6 lb/hr)—535 times the allowable rate.

• Finally, when pressurized to 40 barg (580 psi), which is close to the maximum pressure limit for a “Type A” seal under API 682/ISO21049, the leakage rate is 125 cc/min (4.2 fl.oz/min) or 7,500 g/hr (16.5 lb/hr)—1,300 times the allowable rate.

None of the leakage rates in TABLE 3 complied with the de-sign intent of Section 10.3.1.4.1 in API 682. Moreover, such extreme leakage rates will obviously not meet the reasonable expectations of reliability and safety-focused users.*

ANSI/API RP 754 and safety guidelines of the Baker Panel. The safety-critical implications of the air test vs. actual liquid-leakage discrepancies are staggering. Seals that enter the market based on passing the present API/ISO air test can, in the most extreme case, leak 7,500 g/hr of flammable, toxic or haz-ardous product.

The significance of this issue becomes clearer when we exam-ine another API standard, ANSI/API RP 754. This document was created following the 2005 explosion at the BP refinery in Texas City, Texas. It emphasizes recommendations, which following the Texas City event, were made by the BP U.S. Refineries Indepen-dent Safety Review Panel (the Baker Panel) and the US Chemical Safety Board. The stated aims of ANSI/API RP 754 are to:

• Indicate changes in company or industry performance, to be used to drive continuous improvement in performance

• Perform company-to-company or industry segment-to-segment benchmarking

• Serve as a leading indicator of potential process safety is-sues, which could result in a catastrophic event.

A significant focus of ANSI/API RP 754 is “loss of process containment” (LOPC) events. LOPC events are categorized in

TABLE 1. Key items of API/ISO protocol for mechanical seal air tests

Each sealing section shall be independently pressurized with clean air to a gauge pressure of 0.17 MPa (1.7 bar or 25 psi). The volume of each test setup shall be a maximum of 28 l (1 ft3).

Isolate the test setup from the pressurizing source and maintain the pressure for at least 5 minutes.

Maximum pressure drop during the test shall be 0.014 MPa (0.14 bar or 2 psi).*

TABLE 2. Summary of air-test results

Test description Pressure drop

26.5-l Closed-loop system with virtually airtight mechanical seal and no orifi ce (control test)

Zero pressure drop—Pass

26.5-l Closed-loop no leakage system with 0.008-in. orifi ce added

2-psi pressure drop—Pass

26.5-l System with initially airtight mechanical seal and 0.008-in. orifi ce added

2-psi pressure drop—Pass

1.5-l System with initially airtight mechanical seal and 0.008-in.orifi ce added

22-psi pressure drop—Fail

1.5-l System with virtually airtight mechanical seal and no orifi ce

Zero pressure drop—Pass

TABLE 3. Liquid leakage rates at various pressures with 0.008-in. (0.2-mm) orifi ce

Test pressure Leakage Actual vs. allowable leakage

2 barg (29 psig) 1,500 g/hr 270 * the maximum

7 barg (102 psig) 3,000 g/hr 535 * the maximum

40 barg (580 psig) 7,500 g/hr 1,300 * the maximum


Register by 31 May to SAVE up to 20%Conference registration includes:• Pre-conference tour of Indian Oil’s Panipat Refi nery & Petrochemicals Complex (9 July 2013)*

• Two-day conference program (10-11 July 2013) including keynote addresses, general presentations and panel discussions

• Breakfasts, lunches and refreshment breaks

• Access to exhibition fl oor throughout conference activity *Refi nery tour registration is offered on a fi rst-come, fi rst-served basis. Space is limited.

2013 Conference Fees:

Early Bird Fee (by 31 May) Regular Fee

Individual $945 $1,045

Team of Two $1,700 $1,875

Group of Five $4,250 $4,700

For more information about IRPC 2013, please contact Melissa Smith, Events Director, Gulf Publishing Company, at +1 (713) 520-4475 or [email protected].

❑ Individual ❑ Team of Two ❑ Group of Five

Company _______________________________________________________________________________________

Address ____________________________________________________________________________________________________

City ______________________________________________________________ State _________________________________

ZIP _______________________________________________________________ Country _______________________________

Phone ____________________________________________________________ Fax ___________________________________

Name ____________________________________________________________ Title __________________________________

Signature _________________________________________________________ Date __________________________________

Enclose with payment to: 2 Greenway Plaza, Suite 1020, Houston, Texas 77046 USA

or Register online at HPIRPC.com

EARLY BIRD SPECIAL

Register by 31 May

to Save!

HPIRPC.COM


Early Bird Special Register by 31 Mayto SAVE up to 20%IRPC 2013 Advisory Board Members:

ANDREA AMOROSOVice President, Process Technologyeni - Refi ning & Marketing Division

A. K. ARORADirector GeneralPetroleum Federation of India

JOHN BARICLicensing Technology ManagerShell Global Solutions International B.V.

ANINDYA SUNDAR BASUManaging DirectorChennai Petroleum Corporation Ltd

ERIC BENAZZIMarketing DirectorAxens

CARLOS CABRERAExecutive ChairmanIvanhoe Energy

DR. CHARLES CAMERONHead of Technology, Downstream; VP, Formulated Products TechnologyBP plc

GIACOMO FOSSATAROGeneral ManagerWalter Tosto S.p.A.

DR. MADHUKAR ONKARNATH GARG DirectorCSIR-Indian Institute of Petroleum

RAJKUMAR GHOSHDirector (Refi neries)Indian Oil Corporation Limited (Refi neries Division)

CHAKRAPANY MANOHARANDirector-Refi neryEssar Oil Ltd.

SAMIK MUKHERJEECountry ManagerTechnip India Limited

B. K. NAMDEOExecutive Director,International Trade & SuppliesHindustan Petroleum Corporation Limited

SYAMAL PODDARPresidentPoddar & Associates

A.K. PURWAHAChairman & Managing DirectorEngineers India Limited

GIACOMO RISPOLIExecutive Vice President, Research & Development and Projectseni - Refi ning & Marketing Division

STEPHANY ROMANOWEditorHydrocarbon Processing

DR. AJIT SAPREGroup PresidentResearch and Technology Reliance Technology Group

MICHAEL STOCKLEC Eng FlChemEChief Engineer - Refi ning TechnologyFoster Wheeler

MR.K.VENKATARAMANANChief Executive Offi cer & Managing Director Larsen & Toubro

S VENKATRAMANDirector (Business Development)Gail (India) Limited

IRPC 2013 Sponsor Organizations:

Technip

KBR Technology

GTC Technology

Elliott Group

Carpenteria Corsi s.r.l

Walter Tosto

Axens

Gulf Publishing Company

Petroleum Economist

OilCareers

IndianOil

Petrofed

View the

Agenda at

HPIRPC.COM



four tiers by severity, with the most extreme LOPC events be-ing labeled as “Tier 1” and defined as “LOPC events of greater consequence.” To be categorized as a “Tier 1” hazard, one of two conditions must be met:

1. LOPC results in a lost-time injury, fatality or fire; or2. Any LOPC above a defined size limit, regardless of the

consequence.For an incident to be considered “most serious,” the release

must be classified as an “acute release.” By definition, an acute release must meet a threshold value or quantity limit within one hour. TABLE 4 lists the threshold quantities for a Tier 1 event.

From TABLE 3, a mechanical-seal assembly containing a 0.008-in. (0.2-mm) orifice at a typical pressure (7 barg or 102 psi) was capable of leaking 3,000 g/hr (6.6 lb/hr). Also, it passes the present API air test. If the duty was a TIH Zone A material, leakage rates are so severe that, by ANSI/API RP 754 definitions, the seal would be in a constant state of the most serious report-able “Tier 1” acute leakage. This leakage would be considered a separate new Tier 1 acute LOPC process safety incident each and every hour that the seal was used to contain the process.

Use seals that satisfy common-sense safety demands. When it comes to sealing issues, industry must apply the les-sons from Bhopal and other disasters. It is obvious that a test that can allow up to 1,300 times the intended leakage should be revised. By comparing the test results to API’s own hazard-ous release thresholds, it appears that safety should be sufficient grounds to amend the current standard.

The world’s best petrochemical and oil refining plants are not using API or other industry specifications as stand-alone criteria. The leading users of, say, centrifugal pumps, recognize that speci-fication supplements or amendments are needed. The various cautionary clauses given in the typical introductory statements to different standards are well-known best-practices companies. For example, one of these clauses reads: “The purchaser may desire to modify, delete or amplify sections of this standard.” Another pref-ace is worded: “Standards are not intended to inhibit purchasers or producers from purchasing or producing products made to specifications other than those of API.” Appropriately, the API expressly declines any liability or responsibility for loss or dam-age resulting from the use of its standards, or for the violation of any regulation with which the standards publication may conflict.

Reliability and safety professionals have to manage risks ef-fectively, and standards can indeed be extremely helpful. Test-ing, however, has shown beyond all doubt that not all standards are harmonious and noncontradictory. Invoking an inconsis-tent standard will not sufficiently mitigate risk. When it comes to sealing hazardous, flammable and/or toxic products, engi-neers should understand that just because a seal meets the pres-ent levels that it does not mean this seal will perform safely in service. The recommended action would be for the test loop volume in the API/ISO protocol for mechanical seal air tests to be changed to 1.5 l. This recommended change is reflected in the first point of TABLE 5.

We are encouraged by reliability professionals on several con-tinents who are now taking tangible steps to reduce mechanical-seal failure risks. Also, because of the weakness resulting from the presently allowed large-volume protocol of TABLE 1, at least one major mechanical-seal manufacturer is now voluntarily

subjecting its API 682-qualified seals to the far more stringent “low-volume” air-testing described earlier. TABLE 5 restates the “low-volume” air-testing protocol. Finally, users can reduce the risk of installing flawed mechanical seals and, thereby, increase safety and reliability by specifying, buying and installing only seals with initial liquid leakage rates not exceeding 5.6 g/hr.

NOTES * Footnote: A demonstration video, which includes footage of each stage of testing

detailed in this article, has been released into the public domain. The author active-ly encourages interested readers and safety-focused seal users to review the readily available video: http://www.sealsuccess.com/api-682-mechanical-seals-leaking/.

LITERATURE CITED 1 Bloch, K. and B. Jung, “The Bhopal Disaster: Understanding the impact of unreli-

able machinery,” Hydrocarbon Processing, June 2012.

STEPHEN SHAW, CEng, FIMechE, CMOSH, is a chartered engineer and a Fellow of the Institution of Mechanical Engineers. In addition, he is a Chartered Safety and Health Practitioner. Since 2008, he has been the chairman of AESSEAL plc. He was appointed group engineering director of AES Engineering Ltd., in 2011.

TABLE 4. Threshold quantity limits for a Tier 1 LOPC hazard

Threshold release category Material hazard classification Threshold quantity

1 TIH Zone A materials 5 kg (11 lb)

2 TIH Zone B materials 25 kg (55 lb)

3 TIH Zone C materials 100 kg (220 lb)

4 TIH Zone D materials 200 kg (400 lb)

5 Flammable gases or liquids with initial boiling point ≤ 35°C (95°F) and fl ash point < 23°C (73°F) or other packing group I materials excluding strong acids/bases

500 kg (1,100 lb)

6 Liquids with initial boiling point > 35°C (95°F) and fl ash point < 23°C (73°F) or other packing group II materials, excluding moderate acids/bases

1,000 kg (2,200 lb) or 7 bbl

7 Liquids with fl ash point ≥ 23°C (73°F) and ≤ 60°C (140°F) or liquids with fl ash point > 60°C (140°F) released at a temperature at or above fl ash point or strong acids/bases or other packing group III materials or Division 2.2 nonfl ammable, nontoxic gases (excluding steam, hot condensate, and compressed or liquefi ed air)

2,000 kg (4,400 lb) or 14 bbl

TABLE 5. Proposed key items of API/ISO protocol for mechanical-seal air tests

Each sealing section shall be independently pressurized with clean air to a gauge pressure of 0.17 MPa (1.7 bar or 25 psi). The volume of each test setup shall be a maximum of 1.5 l.

Isolate the test setup from the pressurizing source and maintain the pressure for at least 5 minutes.

Maximum pressure drop during the test shall be 0.014 MPa (0.14 bar or 2 psi).

SEPARATION

ANXIETY?COME TO BOOTH 1005 AT OTCTO LE A RN HOW THIS SK IDDED

SEPA R ATOR CA N HELP!

(936) 788-1000

www.pentairseparations.com

You can relax.Pentair Separation Systems designs and manufactures

advanced technologies for the high performance

separation of solids, liquids and gases.

Dramatically decrease or eliminate contamination like:

Reduce your anxiety.

Increase your efficiency.

SEPARATION SYSTEMS



Special Report Maintenance and Reliability A. GLAUN and J. SHAHDA, GE Oil & Gas,

Avon, Massachusetts

Prevent methane hydrate formation in natural gas valves

Gas flow across a control valve is considered a classic “throttling” process that is defined by energy not being add-ed or extracted from the process gas as it traverses the valve. Therefore, total enthalpy is preserved, entropy increases and the process is thermodynamically irreversible.

The consequences of this process are that many real gases experience a drop in temperature while following the constant enthalpy line as the pressure drops across the valve. This effect was first described by William Thomson and James Joule, and it now bears their names. The Joule-Thomson effect is leveraged in the production of cryogenic fluids such as liquid oxygen, nitrogen and argon, and it is the principle of operation behind most air conditioners and re-frigeration in use today.

Natural gas production, storage and transmission usually take place close to ambient conditions, where a small change in temperature can induce the formation of methane clath-rates (hydrates). Once formed, methane hydrates can block valves, fittings and pipelines. Newer facilities are using higher transmission pressures, causing the temperature inside the valve to approach or drop below 0°C, with the risk of icing on the outside of the valve.

The discussion here focuses on the thermodynamics in-volved and on the requirements for a successful natural gas valve application in which the incidences of hydrate formation and icing of the valve are reduced. Computational fluid dy-namics (CFD) studies are also presented showing the Joule-Thomson effect in a real-world valve application.

What are methane hydrates? Natural gas/methane hy-drates (also known as methane ice) are crystalline water ice-like particles, where methane molecules are trapped inside hy-drogen-bonded water molecules. Under the right conditions of pressure and temperature, these form semi-solid particles that tend to agglomerate, building up inside pipelines, valves and other process equipment.

Why worry about methane hydrates in valves? Hydrate ice particles may clog flow passages in control valves and, in particular, valves with noise attenuation trim (small drilled-hole cages, labyrinth passage stacks, etc.). This sometimes causes a major reduction in the flow across the valve, badly af-fecting system operation. Severe hydrate formation may even clog large passages of the valve body and pipeline.

How do methane hydrates form? Hydrates form in natu-ral gas pipelines when the local fluid temperature drops below the hydrate-formation temperature at a specific pressure. This temperature drop can occur when the natural gas flows through a control valve, or when gas travels through transmission pipe-lines under cold ambient conditions or through any other piece of process equipment where the flow is restricted or accelerated in such an orifice plate. This phenomenon of temperature drop with pressure drop in a real gas is known as the Joule-Thomson effect. Note: Hydrates can form at temperatures well above the freezing point of water (FIG. 1).

Hydrate-formation temperature is difficult to predict and is the subject of many academic papers. Prediction depends on temperature and pressure, water concentration and the compo-sition of the natural gas, where small concentrations of heavy hy-drocarbons and other gases such as O2 , N2 , H2 S and CO2 can af-fect the formation temperature. Software programs are available to help the user predict the formation temperature, but the only way to know for certain is to test a sample of the gas in question.

How is the gas temperature drop calculated? Flow across a control valve is considered a throttling, constant enthalpy (is-

1750

5

10

15

20

25

30

35

40

200 225 250Temperature, K

Methane clathrate, stable

Pres

sure,

MPa

275 300 325

FIG. 1. Stability curve showing that methane hydrate is stable at 0.1 MPa (1 bar) if temperatures are low enough, and that it is stable far above the melting point of ice (H2O) if pressures are high enough. Data courtesy of Lawrence Livermore National Laboratory.



enthalpic) process. This implies that the process occurs over a very short period, making it adiabatic (no heat is lost or gained during the process); enthalpy is preserved, and the process is ir-reversible (i.e., entropy increases and cannot be recovered).

For a real gas flowing through a control valve, this process gives a lower downstream temperature. Additional lowering of the downstream temperature may occur due to high down-stream velocity of the expanded gas (Eq. 1). For natural gas and reasonable downstream velocities of less than 0.3 Mach num-ber (Ma), the velocity terms in Eq. 1 are two orders of magni-tude smaller than the enthalpy and can usually be ignored.

(1) h1 + V1

2

2= h2 + V2

2

2 Where:

h = Specifc enthalpyV = Fluid velocity1, 2 = Upstream and downstream conditions, respectively. Two common methods exist to calculate the temperature

drop of natural gas for a given pressure drop across the valve. The first method is to determine the enthalpy at the inlet pres-

sure and temperature and then to determine the outlet tem-perature at the same enthalpy and outlet pressure. Software programs and web-based calculators can give this data, but the Mollier chart for methane can also be used, assuming an isen-thalpic process in the valve from Eq. 1.

A Mollier chart, at minimum, displays properties of pres-sure, temperature, enthalpy and entropy on one diagram, allow-ing the user to define a state using only two properties and read-ing off the other properties (FIGS. 2–4). By definition, this is an accurate method of determining the downstream temperature; it is only limited by the accuracy of the Mollier chart and by the user’s ability to graphically interpolate the chart. Using soft-ware may be more precise, but the authors believe that a Mollier chart gives the user a visual sense of how the values are chang-ing and leads to a better understanding of the thermodynamics.

After determining the inlet condition on the chart, the user follows the lines of constant enthalpy until the downstream pressure line is reached. The temperature now can be read at this new position. The caveats to this method are that the as-sumption of constant enthalpy is just that—an assumption. In reality, there is some heat transfer across the valve/pipe bound-ary, and the process is never precisely a true throttling process. These “inefficiencies” will result in lower temperatures than the ideal determined above.

The second method is a general rule used in the natural gas industry where, for every 100-psi pressure drop, there is a cor-responding 7°F temperature drop; however, this rule is limited to a maximum valve inlet pressure of 1,000 psi. Using the Mol-lier chart for methane at room temperature, the accuracy of this rule can be evaluated. It varies from 5.5°F/100 psi for in-let pressures of approximately 300 psi, to 6°F/100 psi for inlet pressures of approximately 1,000 psi.

The rule takes into account inefficiencies and is somewhat conservative. However, for high inlet pressures and small pres-sure drops, the rule is very conservative. For example, from a

2703.5 3.7 3.9

Entropy, kJ/kg

Methane throttling process

4.1 4.3 4.5

280 A

1

2B

290

300

310

Constant enthalpyConstant pressu

re, bar

Hydrate formation line

Tem

pera

ture,

K 320

330

340

350200 160 140 120

80

90

100

FIG. 2. Temperature drop inside a single-stage trim valve (Line A) and a multi-stage trim valve (Line B).

FIG. 3. Single-stage contoured plug valve (Line A). FIG. 4. Multi-stage, expanding-area trim valve (Line B).



drop of 1,000 psi to 800 psi, the temperature drop is 4.5°F/100 psi per the Mollier chart.

The temperature downstream of the valve is a concern, but the lowest temperature inside the valve trim must also be calculated. In fact, the pressure and temperature at the trim vena contracta (smallest area of flow in the trim) are usually lower than the pressure and tempera-ture downstream of the valve. The lower temperatures inside the valve are due to sudden accelerations of the gas inside the valve trim and typically are thermody-namically isentropic (reversible) in nature.

These internal pressure drops can be very large for single-orifice valves and less so for multi-stage control valve trims that drop the pressure over a number of controlled pressure-drop stages; they can be visual-ized if the reader follows the constant entropy line on the Mollier chart. Fortunately, this vena contracta low tempera-ture is not a permanent change of state, and the temperature due to this effect recovers after the gas passes through the valve trim, leaving only the vena contracta and adjacent areas cold.

How can hydrate formation be avoided? There are several solutions to reduce the incidence of hydrate formation in valves:

1. Appropriate inlet temperature. The natural gas inlet temperature can be chosen so that, when the pressure drops across the valve, the resulting downstream temperature of the natural gas is always above the hydrate-formation temperature. Gas temperatures are normally determined by the gas field, so external heating of the inlet gas prior to entering the valve may be the only option. This is, however, expensive in terms of heating equipment and fuel costs. The required inlet tempera-ture can be determined using the Mollier chart or the 7°F/100 psi rule, by starting at a known safe outlet temperature and then working backward.

2. Inhibitor injection. Inhibitors can be injected upstream of the control valve to prevent the gas from reaching the hydrate-formation temperature, thereby preventing the formation of hy-drates. The most common inhibitors are methanol and ethylene glycol; these typically can be recovered from the gas and recir-culated. However, inhibitor injection and recovery can be costly.

3. Valve trim design. As noted earlier, even if the valve outlet is above the hydrate-formation temperature, the inter-nal valve trim temperature may not be, and hydrate formation is possible within the valve. If this is the case, then selecting a multi-stage valve that gradually lowers the pressure across the valve trim will help the situation. Note: Trim selection can-not prevent downstream hydrate formation if the downstream temperature is below the hydrate-formation temperature. The Joule-Thomson effect is a “state” condition from upstream to downstream, and changing the valve trim will not affect this.

The red lines in FIG. 2 show the properties of methane as the fluid travels through the control valve from upstream (1) to downstream (2). The long, dashed red line labeled “A” repre-sents a single-stage control valve where the temperature drops below the hydrate-formation line (blue line), making it pos-sible for hydrates to form inside the trim. The dotted red line labeled “B” represents a multi-stage control valve where the temperature does not drop below the hydrate-formation line, thus preventing hydrates from forming inside the trim.

Valve icing. Under high pressure-drop conditions, the outlet temperature in the valve may fall below the freezing point of water. This may not cause hydrate formation inside the valve

because inhibitors such as monoethylene glycol (MEG) can be used with gas at −10°C. Even so, condensation and freezing on the outside of the valve body and pipeline can have serious ef-fects. For example, coastal gas fields on the Saudi Arabian pen-insula are notoriously humid and prone to ice buildup.

Extremely thick layers of ice can build up, preventing access to the valve body or pipe wall. These layers of ice can add sig-nificant weight to the valve and pipeline, with the possibility of structural and/or vibration problems. The valve bonnet may become iced, thus seriously impacting the valve stem packing and raising the potential for leakage.

A real-world problem. A natural gas producer was flowing gas through a control valve with the following winter conditions:

• Upstream: 975 psia, at 57°F• Downstream: 180 psia, with icing on the valve and pipe.The icing was unacceptable to the plant operator, and the

only line heaters available were rated at 350 psia and could not be used to heat the inlet gas. For the purposes of this example, the natural gas is assumed to be methane.

An isenthalpic analysis showed that the downstream tempera-ture reaches 14°F (FIG. 5, points 1–2). Using a gas industry gen-eral rule, the downstream temperature could reach 1.3°F (FIG. 5, points 1–3). The end user required that the downstream temper-ature be no less than 40°F to prevent icing and hydrate formation.

Hydrate-formation temperature is difficult

to predict and is the subject of many

academic papers. Software programs are

available to help the user predict the

formation temperature, but the only way

to know for certain is to test a sample

of the gas in question.

1 5

423

6

Constant temp.

Min. gastemp. 57°FInlet pressure

Outlet pressure

40°F

20°F60°F

80°F

100°F

0°F

Heat input 59 kJ/kg

Original–isenthalpicOriginal–7°F/100 psiSolution 1–isenthalpicSolution 1–7°F/100 psi

0

200

400

600

800

1,000

1,200

770 790 810 830 850 870 890Enthalpy, kJ/kg

Pres

sure,

psi

FIG. 5. Heating required at inlet pressure to keep outlet temperature above 40°F.



Solution 1: Heat gas at the valve inlet. A common solution to this type of icing problem is to use pipeline heaters just up-stream of the valve. The fuel for the heaters is usually the flow-ing natural gas itself. However, this solution is costly in terms of lost gas and the expense of high-pressure heaters.

Referring to FIG. 5 and using the 40°F minimum outlet tem-perature requirement (dashed blue line), point 4 can be located and the temperature can be back-calculated to maintain 40°F at the outlet of the valve. The gas inlet temperature should be above 95°F (point 6) to avoid falling below 40°F at the outlet.

Using an isenthalpic analysis, the inlet temperature should be above 84°F (point 5) to avoid falling below 40°F at the outlet. The difference between points 5 and 6 is quite substantial, and, as discussed earlier, the authors believe that, for high pressures, the 7°F/100-psi rule is overly conservative. The isenthalpic anal-ysis is ideal for this measurement; the true value lies somewhere in between.

Note: The addition of hydrate inhibitors might lower the end user’s specification of 40°F minimum temperature at the outlet of the valve. In this case, point 4 would move to a lower value to the left and the analysis would be repeated, thereby lowering the min-imum required inlet temperature to prevent hydrate formation.

The rate of energy input required to heat the gas can be read directly from FIG. 5 by subtracting the enthalpy at point 6 from the enthalpy at point 1. If this value is multiplied by the mass flowrate in kg/s, then the answer is the rate of energy input in kJ/s or kW.

As mentioned before, this analysis is independent of the type of trim in the valve. If the analysis shows that the tempera-ture at the outlet is low enough to form hydrates, then changing to a multi-turn or multi-stage trim will not alter the conditions at the outlet. A multi-stage valve will, however, limit very low temperatures inside the valve trim.

Solution 2: Use available low-pressure heaters. The first option considered was to save the customer from having to buy new equipment by using the existing, 350 psia-rated line heat-ers (FIG. 6). This method required staging the pressure drop by placing another valve in the line. The first pressure drop oc-curred from 975 psia to the heater maximum pressure of 350 psia, and then down to the outlet pressure of 180 psia.

In the methodology of this solution, the outlet drop should not fall below 40°F, which allows point 4 to be located. An is-

enthalpic analysis is used to back up to the heater pressure of 350 psia, which gives points 3 and 5, respectively. Point 2 is located on the 40°F minimum line, and an isenthalpic analysis is used to back up to the inlet pressure of 975 psia, giving points 1 and 6, respectively. The heat input is calculated from points 2–5. The addition of heat at 350°F reduces the minimum inlet temperature to 84°F from 95°F, with no heat addition.

Note: As with solution 1, the addition of hydrate inhibitors might lower the end user’s specification of 40°F minimum at the outlet of the valve. In this case, point 4 would move to a lower value and the analysis would be repeated, thereby low-ering the minimum required inlet temperature to prevent hy-drate formation.

Solution 3: Apply new low-pressure heaters. If the an-swers from the first two solutions are inadequate, the next step is to examine the lowest-pressure-rated line heaters that can be used and still operate year-round at the minimum inlet tem-perature of 57°F. This requires a slightly different methodology than that used previously.

Referring to FIG. 7 and starting at the minimum inlet tem-perature at point 1, the pressure must then be determined for when 40°F is reached. This gives point 6, which is at 732 psia. Knowing that the endpoint is point 4, one can work backwards, using isenthalpic analysis, to arrive at point 5. The enthalpy difference between points 5 and 6 is the resulting heat input required. Comparing the result of solution 3 to solution 1, a small reduction in heat input is required. Note that the heat in-put found when using the general rule is identical to that found when using the isenthalpic analysis.

At this point, it becomes a question for the end user of eco-nomics and complexity. Solution 1 appears to be less complex, since it requires only one control valve; however, a large pres-sure drop across one valve results in a severe service applica-tion with low internal valve trim temperatures, possibly requir-ing an expensive multi-turn or multi-stage valve. High-rated pressure-line heaters also must be purchased, and significant heat must be added to the upstream gas.

Solution 2 does not appear to be useful since an additional valve would need to be added to the line to accommodate the pressure drop from 350 psia to the outlet pressure of 180 psia, and the system would not be able to run unless the ambient

2

1 6

3

4

5

Constant temp.


Outlet pressure

Heater max. pressure

40°F

20°F

60°F

80°F

100°F

0°F

0

200

400

600

800

1,000

1,200

770 790 810 830 850 870 890Enthalpy, kJ/kg

Pres

sure,

psi

Heat input 16 kJ/kg

FIG. 6. Minimum inlet temperature when using the existing low-pressure line heaters.

2

1

36

4

5Constant temp.


732 psi

Outlet pressure

40°F

20°F

60°F

80°F

100°F

0°F

0

200

400

600

800

1,000

1,200

770 790 810 830 850 870 890Enthalpy, kJ/kg

Pres

sure,

psi

Heat input, 56 kJ/kg

FIG. 7. Heat input at lowest line heater pressure for preventing hydrates at minimum inlet temperature.



(inlet) temperature reached 84°F. However, there are some geographical locations where this might not be such a burden.

Solution 3 requires the complexity of an additional valve, but the pressure drop is broken up into two reasonable steps, resulting in two less severe applications and warmer internal valve trim temperatures. Lower-rated pressure-line heaters would need to be purchased, and significant heat would need to be added to the upstream gas.

Numerical analysis for valve trim temperature. CFD can be used to model the flow through the trim of the valve. An accurate CFD analysis to capture Joule-Thomson effects is only possible if advanced real gas formulations are used. A real gas model takes into account non-ideal compressibility effects, whereas an ideal gas CFD analysis will only predict localized drops in temperature resulting from increases in ve-locity and reductions in local pressure due to the acceleration of the fluid as it negotiates turns in the valve trim.

The proprietary CFD program used in this case has a real gas model that uses the Redlich-Kwong formulation to pre-dict the fluid properties, taking into account the non-ideal compressibility of the working fluid. The program predicted 54°F at the outlet of the trim, which compares favorably to an isenthalpic analysis using a Mollier chart, which predicts 54.5°F.

FIG. 8 shows a representative 22-turn trim (pictured is a half-symmetry model of one flow channel) with an inlet

pressure of 975 psia and an outlet pressure of 180 psia. The results show the even, gradual, staged pressure drop through the valve trim. This style of trim serves two main purposes: One is to lower the outlet jet Ma, producing a quiet valve, and the other is to reduce the temperature drop inside the valve trim to minimize hydrate formation and icing.

FIG. 9 shows the temperature results. The plot clearly shows the Joule-Thomson effect of a permanent temperature drop from inlet to outlet. It also shows areas inside the valve trim

FIG. 8. CFD pressure plot of a representative 22-turn, multi-stage valve trim.

+1 [email protected]©T.F. Hudgins, Incorporated. All rights reserved.

Oil Mist Lubrication. Worldwide.

Need oil mist lubrication expertise?� New Systems� Accessories� Installation� Repairs� Support

For any oil mist application –in any location – we’ve got you covered.




where the temperature can drop below the outlet temperature, although areas are localized and the temperature recovers.

Even accepted global standards for valve sizing, such as IEC 60534-2-1, do not take this real gas effect into account and base the sizing exclusively on upstream temperature, assuming an ideal gas where interstage and downstream temperature equals the upstream temperature. (Note: IEC 60534-2-1 does warn that compressibility of real gases should be taken into account if an accurate upstream density is to be calculated.)

IEC control valve noise prediction standard 60543-8-3 explicitly states in its scope statement that ideal gas laws are assumed, and it uses the upstream temperature to determine downstream density, velocity and Ma. For this specific prob-lem, the downstream velocity can be under-predicted by 8%.

Takeaway. Hydrate formation and icing in natural gas pipe-lines and valves can be greatly reduced or even prevented en-tirely if a detailed study of the thermodynamics of the system is undertaken. An intimate knowledge of the process gas is es-sential so that properties, such as hydrate-formation tempera-ture, can be accurately determined. Also, using real gas analysis, internal valve trim temperatures can be calculated, leading to a better understanding of the type of valve trim required to in-hibit hydrate formation.

ASHER GLAUN is a senior engineer and technologist for Masoneilan Control Valves at GE Oil & Gas. He has worked in the control valve industry for over 12 years. Prior to his work with GE Oil & Gas, Mr. Glaun was employed for 11 years at Bird Machine Co. in the design of high-speed centrifuges. His work at GE involves leading new technology development specializing in fluid dynamics, CFD, structural analysis/FEA and valve acoustics.

Mr. Glaun graduated with a BSc degree in mechanical engineering from the University of Cape Town, South Africa, and he obtained an MS degree in mechanical engineering from Northeastern University in Boston, Massachusetts.

JOSEPH SHAHDA is a senior applications engineer for Masoneilan Control Valves at GE Oil & Gas. He has over 16 years of experience in the control valve industry, with a focus on applications engineering and delivering control valves solutions to customers worldwide. Mr. Shahda holds an MS degree in mechanical engineering from Northeastern University in Boston, Massachusetts.

A01

120E

N

Partner with the Best

With over 50 independent subsidiar-ies and more than 220 engineering and sales offi ces spread across the world, SAMSON ensures the safety and environmental compatibility of your plants on any continent.

To offer the full range of high-quality control equipment used in industrial processes, SAMSON has brought together highly specialized compa-nies to form the SAMSON GROUP.

SAMSON AG · MESS- UND REGELTECHNIK Weismüllerstraße 360314 Frankfurt am Main · GermanyPhone: +49 69 4009-0 · Fax: +49 69 4009-1507 E-mail: [email protected] · www.samson.deSAMSON GROUP · www.samsongroup.net

Visit us atOTC Houston 6 to 9 May 2013Hall B, Booth 4717-5

FIG. 9. CFD temperature plot of a representative 22-turn, multi-stage valve trim. A real gas solver formulation allows for the solving of Joule-Thomson effects.


Bringing energy and the environment into harmony.®

Visit us at OTC2013, Reliant Arena Hall, Booth 9659, Houston, Texas, May 6-9

RELIABILITY ISSUES NEVER SLEEP. WE MAKE SURE YOU CAN.

At Dresser-Rand, we know any unscheduled interruption of a client’s downstream operation is simply unacceptable. That’s why we make engineering reliability a primary focus. From one of the world’s largest hydrogen compressors (pictured at right) to our single-stage steam turbines, Dresser-Rand downstream solutions are easy to maintain. But for us, it’s not just about selling superior products, it’s about providing peace of mind as well. So, when you do need help, rest assured we’re there for you.

» www.dresser-rand.com

The Americas: (Int’l +1) 713-354-6100 / EMEA: (Int’l +33) 2-35-25-5225Asia-Pacifi c: (Int’l +60) 3-2093-6633 / [email protected]

Compressors–Turbo & Recip / Steam Turbines / Gas Turbines / Engines / Control Systems / Expanders



Special Report Maintenance and Reliability H. P. BLOCH, PE, Consulting Engineer,

Westminster, Colorado

Consider both actual and virtual spare parts inventory

Globally, hydrocarbon processing industry (HPI) facilities operate hundreds of thousands of fluid machinery assets. The various service-specific internal components of these assets are supported by many millions of spare parts, all of which must be properly preserved and catalogued.

Operating companies and consultants often consider de-veloping systems to unify or standardize spares maintained in plant stores or inventories. Standardization initiatives can provide economic benefits. Proper and unified part identifica-tion efforts usually precede standardization of the spare parts inventory. Both identification and standardization can help reduce the number of spare part units held in inventory and reduce plant costs. Also, access to spares will be facilitated in instances where central part depots are involved.

Dual-purpose units. Some spares could be manufactured with “multi-purposes” in mind. Multi-purpose parts could be produced with oversize dimensions, which, upon finish-machining, will suit the specific requirements of a particular machine. Certain bore dimensions (for example, in coupling and impeller hubs) could be stocked with undersized and prebored concentricity-accurate pilot holes. On an as-needed basis, these bores could be finish-machined to the precise dimension on site. A single maximum diameter impeller, as shown in FIG. 1, could be the standardized spare part for several pumps. This impeller could be trimmed for use in a variety of pumps presently operating at a particular site and used in dif-ferent pumping services. Similarly, an oversized shaft could be placed in the storeroom for future finish-machining and used on a range of similar pumps or other fluid machines.

Consider everything. The concepts of shared capital spares and/or requiring the vendor to stock spare parts are not new.1 Both concepts—shared spares and vendor stock-ing—warrant more investigation. Major turbo-machinery (fluid machinery) rotors can be jointly owned by a group of potential users, call it an “ownership pool.” The financial details for being able to use and replace co-owned spares should be well defined in advance as part of the shared own-ership pool. Vendor stocking arrangements place the vendor under contractual obligation to have certain machines or parts available on very short notice. In return, the potential user of these parts or machines accepts the contractual ob-ligation to purchase vendor-stocked assets at a predefined premium cost.

Case for stocking only upgraded parts. Many spare parts kept in storage at HPI facilities belong to fluid machines that are not optimized in terms of hydraulic efficiency, permissible range of operation, or component reliability. For decades, capturing efficiency gains and avoiding equipment failures through proper upgrading have been the priority concerns of industry. Moreover, maintenance-cost reductions have been achieved through selec-tive upgrading of existing parts and machinery. This upgrading is usually planned by the user or equipment owner, and such ef-forts receive considerable input from competent vendors. Cost justifications are developed jointly and, if the results warrant, are implemented during the next repair or scheduled downtime.

Whenever upgrading is done, the issue of spare parts adequa-cy should be investigated. Fortunately, as of 2013, there are truly advanced options that go well beyond simply buying new physi-cal spare parts. These options should be investigated and evalu-ated by forward-thinking of the plant and equipment operation. Such exercises would engage assistance from design engineers or consulting contractors. Before standardizing old, existing parts, consider if these parts will be used again. Gain input on this question before moving forward. Understand that the future of the spare parts business is a new direction for HPI facilities.

Prioritize candidate machines for upgrading. A facil-ity’s or corporation’s list of failure frequencies is one primary way to determine where future upgrading will be most cost-effective. These failure records often reveal weak links or repeat failures that should be examined in greater detail. Risk-prone components deserve to be upgraded whenever life-cycle costs (LCCs) are favorable.

There are many reasons why upgraded components for fluid machinery will often differ from the old original parts. Modern or state-of-the-art designs may incorporate advanced metallurgies or adjusted hardness of existing metallurgical compositions. Hardness adjustments are often based on wear amounts observed in the fluid environment of a particular pro-cess unit. Actual hydraulic and wear-related performance data are needed to impart value to a new spare part. These data are supplied either by site engineers or a designated, experienced, consulting entity.

There can be desirable changes to a more favorable duc-tility of ferrous parts, with upgrade opportunities using advanced stainless steel compositions. These changes are particularly advantageous in oil sand projects and other ap-



plications where abrasive wear is prevalent. Upgrading to a more suitable impeller (FIG. 1) has sometimes allowed shut-ting down one pump in fluid loops which previously had two pumps operating in parallel.

Fluid machines older than 20 years (especially, process pumps) are in the top tier of candidates for upgrades. One proceeds by studying such operating parameters as head vs. flow at best efficiency performance (BEP), avoiding excessive loads on bearings; transient conditions at startup and shut-down, and the establishment of best sealing alternatives. These projects will include flush plans selected for seal reliability and for better power efficiency. The work requires involvement of experienced personnel. Whoever is given the upgrade task should receive the full support of plant engineering personnel.

In well-managed corporations with extensive computerized maintenance management systems (CMMS), one would first investigate the extent to which relevant failure data are avail-

able from these systems. As-built data sheets are sometimes available from a user’s own central engineering group. In all instances, the as-running or operational performance of fluid machines must be ascertained before standardization studies can be expected to yield maximum benefit.

Partner with advanced reverse-engineering capabilities. Competent upgrade vendors view every repair event as an op-portunity to upgrade. These vendors should be considered for partnership discussions. The best upgrade vendors use con-tour mapping and measuring machines for reverse-engineering of existing parts. Their expertise is important. One reference claimed to have found 14% of in-stock spares at a petrochemi-cal plant to be unusable.1 Not to be outdone, the machine shop superintendent of a major US oil refinery claimed that 30% of the spare parts at his location were incorrectly dimensioned or simply unserviceable for a variety of reasons.

When working with a competent upgrade company’s nearest vendor shop, the shop, regardless of location, will be supported by the upgrade company’s home engineering organization. Also, the best-equipped and most promising upgrade providers have high-ly modern fluid-machinery testing facilities, as shown in FIG. 2.2

The local shop will usually start its work with automated contour mapping of the parts to be replicated or improved. All dimensions are stored in a computer. However, based on the possibility of dealing with deficient parts, the upgrade company will use its considerable experience and judgment to monitor the accuracy of these parts.

Simply reverse-engineering without further establishing the correctness of such efforts could replicate certain vulnerabili-ties. Impeller or blade contour adjustment studies are made on some parts and must be catalogued. Available “tweaking op-tions” are sometimes evident only to the best upgrade provid-ers. The LCC assessment of upgraded parts becomes a key in-gredient of further decision-making.

While equipment owners may wish to select a particular vendor as the single-source supplier, this upgrade provider will have to demonstrate full competence in understanding many different machines, models, brands and configurations. But this is not where the provider’s demonstration of competence ends; it is, rather, where it begins.

The 21st century environment demands technology-opti-mized methods and procedures. In this environment, an up-grade provider must demonstrate the ability to create a virtual spare parts storehouse. Actual physical spare inventories are used in limited cases; these are situations where they still make economic sense.

Advantages with a virtual inventory. In the future, spares parts will not be sitting in an HPI facility’s warehouse. The fu-ture is in “make it as you need it.” Major multinational equipment upgrade companies use best-available technology throughout.

In general, these upgrade experts “paint” existing parts with a scanner. A three-dimensional (3D) model is constructed from the scanned data. The work flow goes from an electronic library (a virtual inventory) to a printed sand-mold process in which no pattern making is involved. The entire process can take only a fraction of the total time that it would have needed for conventional methods.

D' D

FIG. 1. A single impeller with diameter, D, could be site-trimmed to the more exact diameter, Dˇ. Doing so would allow it to suit the needs of a particular operating point or pumping service. Note: The side plates will remain at the maximum diameter for maximum preservation of energy (high efficiency).

FIG. 2. Partial view of a modern pump test stand with variable speed drives. Source: Hydro, Inc., Chicago, Illinois.



The actual process uses 3D printer technology to print a sand mold directly onto a print bed. No drawings are created, and this true state-of-the-art process eliminates the need for hard tooling. The process also supports geometries that were once restricted to production-quantity investment castings. Even complex cores and undercuts are now readily obtained with modern 3D printer technology. The process is obviously ideal for spare parts that, by definition, are needed in small quantities. Surface finishes and mechanical properties are vir-tually identical to traditional sand castings, and tolerances are superior to many traditional processes. Sand molds can often be produced in less than a single week. So, if a pump designated as your “A” pump needed repairs after 20 months of operation—the mean-time between failure (MTBF)—and it took 20 days before it could be properly repaired, realize that there is a 97% probability that the “B” pump will operate flawlessly during those 20 days. Chances are that management is willing to take the 3% probability or risk an outage that exists for the parallel or “spare” pump. A rapid repair with an old part would also take time, and result in “old,” inefficient or failure-prone operation with parts that really deserved to be upgraded.

Final considerations. Consider fluid-machinery capital spare pooling and vendor stocking programs. Both are con-siderations that deal with standardization and cost savings. Spare-parts cataloguing should, initially, concentrate on high-value or repeat-failing parts; upgrading should be part of the

standardization efforts. Electronic libraries and virtual elec-tronic inventories exist today at some forward-looking loca-tions. They are the future.

While creating a virtual parts inventory of upgraded parts can be a massive challenge, accomplishing the task will be worth it. The ultimate savings will be huge. Enlist a compe-tent equipment upgrade company, one with international presence. Pick a group with the most modern tools and with equipment testing capability.2

Allocate funds for a demonstration project. Work with a consulting company; with their input, let the upgrade com-pany produce the upgraded components wherever cost justifi-cations can be shown. Work toward a target return on invest-ment (ROI). If only a somewhat lesser ROI is achieved, then there will still be improvement.

So, to get things off to a manageable start, identify possibly 10 repeat-failure or low-MTBF process pumps. Ask the con-sulting company and parts provider to actually produce the up-graded components wherever cost justifications can be shown. Report on tangible achievements and then expand your base to include more assets. Do not make the mistake of making stan-dardization an unwieldy open-ended task. Closely scrutinize the process, as illustrated in FIG. 3, and, if needed, use it to over-come an operations department’s objections. The operations department may be averse to running a process pump without having a spare (or standby) pump sitting on the adjacent foun-dation. Explain to the operations staff that there may be incre-

PROVEN EXPERIENCE. TRUSTED RESULTS.WWW.CUDD.COM

Managing turnarounds on time and on budget can present many challenges. Cudd Energy Services helps you meet these challenges head on. Our

emission rating helps reduce emissions.

EFFICIENCY MATTERS.RESTORE ONLINE STATUS QUICKLY AND SAFELY.


72


mental time needed for components to make their way from an electronic library to becoming physical parts. However, de-pending on the MTBF and days of unavailability because the spare pump is out for repair, the probability of the presently operating pump failing in that time may be acceptably low.

The next phase in the standardization and upgrading process, involves quantifying the deliverables. Perhaps an experienced consulting company, in conjunction with the upgrade provider could commit to a full review of a given number of machines per month. Throughout the process of physically upgrading com-ponents or producing a virtual inventory, remember that the quality of any job or task will only be as good as the people in-volved. We advocate involving experienced and well-motivated individuals in the task of identifying upgrade candidates. From there, consider progressing to a virtual library for all assets in the facility. As with everything else, you get what you pay for. The ultimate results of these endeavors may not be immediately obvious, but they will become quite evident in time.

LITERATURE CITED 1 Bloch, H. P., Improving Machinery Reliability, Third Edition, Gulf Publishing Co.,

Houston, 1998. 2 Bihler, K., D. Dominiak, B. Keith and J. Johnson, “Apply new pump software to

test performance,” Hydrocarbon Processing, October 2012, pp. 91–96. 3 “HP In Reliability,” Hydrocarbon Processing, August 1992, p. 25.

HEINZ P. BLOCH resides in Westminster, Colorado. His professional career began in 1962 and included long-term assignments as Exxon Chemical’s regional machinery specialist for the US. He has authored over 520 publications, among them 18 comprehensive books on practical machinery management, failure analysis, failure avoidance, compressors, steam turbines, pumps, oil-mist lubrication and practical

lubrication for industry. Mr. Bloch holds BS and MS degrees in mechanical engineering. He is an ASME Life Fellow and maintains registration as a Professional Engineer in New Jersey and Texas.

11

2

3

45

7

10

15

20

50

100

3 4 7 10Mean-time-between-repairs (MTBR), months

Repa

ir tim

e or d

ays u

nava

ilable

R = 95%R = 97%

R = 98.5%

R = 99%

R = 99.5%

R = 99.7%

R = 98%

20 50 10040301552

FIG. 3. The probability that a “spare” pump with an MTBR of 15.5 months will operate flawlessly in the 7 days it takes to repair the “main” pump will be 98.5%.3

ARCA Flow Group worldwide:

• Competence in valves, pumps & cryogenics• Subsidiaries and partners in Switzerland, the Netherlands, India, P.R. China, Korea and Mexico!

ECOTROL® control valveAdvantages, that should not be kept quiet!

● High reliability guaranteed by precision manufacturing processes and quality control● Emission control and leakage conforming to the highest international standards● Tubeless, integrated mounting of positioners acc. to VDI 3847● Minimal life cycle cost● A range of awarded patents

Take advantage of the most technically innovative control valve in a generation, up to DN 400 (16")!

ARCA Regler GmbH, D-47913 TönisvorstPhone +49-2156-7709-0, Fax …-55, [email protected]

www.arca-valve.com

The «State of the Art» solution!

Request Technical Information or a Quotation:

The

Qui

et W

ork-

Hor

se!

ACHEMASIA

Beijing

13.–16. May 2013

Booth E18


Farris Total Pressure Relief Solutions Include:Pressure Relief Valves

PSM Engineering /Audit Service & Software

Asset Management Services & Repairs

The Total Solution.It’s simple. Safety relief valves and services to support your facility’s entire

lifecycle. Pressure relief systems designed right from the start. Pressure relief

valves that help your plant operate safely. Global valve maintenance and repair

support. And PSM software and audit services to ensure your plant operates the

way it was designed to during an overpressure event.

Total Pressure Relief Solutions.

We invite you to see all that we can do for you. http://farris.cwfc.com or +1 440-838-7690

Overpressure risk never goes away.

That’s why we offer total pressure relief solutions.



Special Report Maintenance and Reliability M. G. CHOUDHURY, A. KULKARNI and D. KORANGA

Reliance Industries, Mumbai, Maharashtra, India

Improve design for pump suction nozzles

Several analysis methods exist to determine the effective-ness of the pump suction piping system’s first adjustable sup-port and how that impacts nozzle loading. Parameters for consideration include how to model the trunnion dummy leg, determining the thermal load for the dummy leg and accurately predicting the first adjustable support base plate’s stiffness.

DIAGNOSING THE PROBLEMThe pump piping system is a common component of most

hydrocarbon processing complexes. It is essential that this sys-tem is analyzed correctly, so that future problems can be mini-mized. Within this system, the first support adjacent to the nozzle plays a vital role during analysis. This support is often an adjustable type that helps in maintaining the pump shaft axis in line with the nozzle axis.

A popular analysis program uses the pipe centerline mod-eling approach. Generally, an ambient temperature pipe trun-nion from the pump centerline to the trunnion support base plate is modeled during analysis. This approach is sometimes criticized, as the trunnion pipe is welded at the main pipe bot-tom and the diametrical growth of the pipe is ignored. To cir-cumvent this criticism, some analysts advocate a more refined analysis using the element from the suction pipe centerline to the bottom of the suction pipe, where the trunnion has been welded as a rigid element, with a temperature the same as that of the suction pipe. Then a trunnion with ambient temperature and without fluid density and pressure is modeled.

A commonly used analysis technique is the centerline anal-ysis method. In this method, the dummy is modeled from the straight pipe neutral or at the mid-node of a three-node elbow, as shown in FIG. 1. The trunnion itself is assumed to be at ambi-ent pressure and temperature. This analysis does not address the issue of pipe diametrical growth at elevated temperatures. The rigid-element analysis addresses this issue.

To address the diametrical pipe growth issue, rigid-element analysis is used. In method, a rigid element is modeled from the pipe centerline to its periphery. Then the dummy with ambient temperature and pressure properties is attached to it (FIG. 2). The rigid element takes care of the radial expansion of the parent pipe. Due to this additional radial expansion, there will be a substantial rise in vertical pipe displacement, which increases nozzle loading.

CASE STUDYThis sample scenario offers an opportunity to see the di-

rect effect of both techniques. Consider a side-suction pump system at a 340°C design temperature, as shown in FIG. 3. This system is then analyzed using a recent edition of popular pipe stress analysis software.

Centerline modeling in straight line Centerline modeling at elbow

FIG. 1. Normal centerline analysis method.

Rigid element

FIG. 2. Rigid-element analysis method.

FIG. 3. A side-suction pump system at a 340°C design temperature.



Various cases in the software are then formed:a) Normal centerline method: For this case, only a dummy

is modeled in front of the nozzle with a sliding base support, with FIG. 4 showing µ = 0.1.

b) Rigid element method: in this instance, the rigid ele-ment is modeled from the pipe center to the parent pipe’s outer periphery (see Node A to Node B in FIG. 5) while maintaining the same conditions as the parent pipe. Then a dummy is at-tached at Node B, with the base support still sliding.

TABLE 1 summarizes the resulting nozzle loads from both analysis methods.

TABLE 1 demonstrates that the loads increase tremendously during the actual simulation of the pipe’s diametrical expan-sion. This can be attributed to substantial nozzle displacement in the vertical direction, which increases the vertical loads.

There are many ways to counter this vertical loading; the obvious option is changing the line’s routing. But this and oth-er methods increase the overall cost to the system. However, one other way can decrease the loading on the nozzle.

The first adjustable support was closely examined for sup-port stiffness implications on the nozzle loading. Normally in software analysis, default stiffness in the program is used. FIG. 6 shows a typical adjustable support.

FIG. 4. Normal centerline method.

FIG. 5. Rigid-element method.

Center line

Guide gap

Guide gap

Pipe trunnion

Plan

Elevation

Center line

FIG. 6. The trunnion support rests on a thin stub-base plate, which is attached to the ground with the help of bolts and a series of plates.

TABLE 1. Nozzle loads tabulated

FX, N FY, N FZ, N MX, N-m MY, N-m MZ, N-m DX, mm DY, mm DZ, mm

Normal centerline method 4,564 2,256 6,021 –987 –4,309 4,172 0 0 –1.936

Rigid element method 4,563 73,437 12,948 77,036 –4,308 6,507 0 0.008 –1.936

TABLE 2. The results show reduction in the vertical loads

FX, N FY, N FZ, N MX, N-m MY, N-m MZ, N-m DX, mm DY, mm DZ, mm

Normal centerline method 4,564 2,256 6,021 –987 –4,309 4,172 0 0 –1.936

Rigid–element method 4,563 73,437 12,948 77,036 –4,308 6,507 0 0.008 –1.936

Rigid–element method with actual support stiff ness

46,611 15,641 7,973 14,715 –4,372 4,562 0 0.002 –1.936


�77

As FIG. 2 illustrates, the trunnion support rests on a thin stub-base plate that is attached to the ground with bolts and a series of stainless steel (SS), polytetrafluoroethylene (PTFE) and carbon steel (CS) plates.

The base plate material will be the same as the trun-nion material. The PTFE plate inserted between the SS and CS plates will reduce the friction loads as its fric-tion co-efficient is 0.1.

The most common assumption made during this analysis is that the first adjustable support’s stiffness is the stiffness default value. In the case of this software analysis, it is 1.0E12 N/mm. This high value of stiff-ness corresponds to a rigid base, like solid ground or some concrete structures. Since this scenario features a support that is effectively resting on a plate 10-mm thick, the stiffness will be lower.

STIFFNESS CALCULATIONTo predict the effects of the first support stiffness on the

nozzle loads, the support stiffness must be calculated accurate-ly. The stub base plate’s vertical displacement must be accurate to calculate the plate’s vertical stiffness.

The simulated vertical displacement results are shown in FIG. 7. For simplification, the scenario assumed that the vertical bolt stiffness is sufficiently high as to assume it to be acting as a vertical restraint. The plate dimensions considered were 250 mm � 250 mm � 10 mm, with a 1,000-N external load ap-plied on the trunnion periphery. From the stiffness equation, F = KX, for an applied load of 1,000 N, and vertical displacement of .0031512 mm, the vertical (Y) base plate stiffness is calcu-lated as 31,7339 N/mm.

UTILIZATIONBy utilizing the base support stiffness, a more accurate

analysis can be performed. TABLE 2 lists the results of all three analysis techniques.

TABLE 2 shows that the vertical loads are considerably re-duced. This is due to the fact that the vertical displacement was reduced by utilizing the calculated support stiffness.

By predicting the accurate support stiffness, more accurate analysis of the pump suction piping is possible. Even after us-ing the pipe element radial growth, the nozzle loads have not in-

creased. This analysis type approach may be adopted when it is necessary to do a rigorous analysis for a critical pump nozzle load.

FINAL ANALYSISFor high-temperature lines, diametrical pipe growth plays an

important role in assessing the nozzle loading on a pump. When incorporated in a pump system analysis, this thermal loading will greatly increase the loading on the nozzle. To minimize the overall alteration of the system, it is advisable to use the actual support stiffness of the first adjustable support, thus reducing the overall cost to the system.

FIG. 7. Simulated vertical displacement results.

For high-temperature lines, diametrical

pipe growth plays an important role in

assessing the nozzle loading on a pump.

When incorporated in a pump system

analysis, this thermal loading will greatly

increase the loading on the nozzle.


| Bonus Report

PETROCHEMICAL DEVELOPMENTSThe petrochemical industry is a complex global business.

Feedstocks dominate total production costs for hydrocarbon-

based products. Consumer demand for petrochemical-based

end products will sustain an annual growth rate of 4%

through 2018. Developing nations are the drivers for the

expanding petrochemicals industry, and they will account

for 70% of the global ethylene market.

Ariel view of Braskem’s UNIB 2 RS cracker at Triunfo

located in the state of Rio Grande Do Sul, Brazil. Besides

the cracker, Braskem has its Innovation and Technology

Center and the world’s first green ethylene plant, part

of the green polyethylene (PE) production, along with

three PE and two polypropylene industrial units located

in the Triunfo Petrochemical Complex. Photo courtesy of

Braskem America.


Bonus Report Petrochemical Developments J. FU, C. ZHAO and Q. XU, Lamar University,

Beaumont, Texas

Consider novel CGC and front-end depropanizer system for olefins production

In many liquid-cracking ethylene facili-ties, the multistage charge-gas compressor (CGC) is used with a front-end depro-panizer (DeC3) process. The optimal de-sign and operation of an integrated CGC-DeC3 system can provide energy savings and other benefits. In the presented case study, an innovative design for an inte-grated CGC-DeC3 system introduces part of the heavier condensate from the early compression stages directly into the high-pressure (HP) depropanizer. Result: The compression work and CGC stripper loading are reduced. This design also low-ers heat duty for part of the heat-exchang-er network. Using rigorous simulation, this new conceptual design can offer sig-nificant energy-saving potential.

BACKGROUNDSteam cracking is the dominant ole-

fin-production method. It includes sev-eral basic processing processes: cracking feedstocks, quenching charge-gas ef-fluents, compressing cracked gas com-bined with drying (deacidification and dehydration), chilling the charge gas, hydrotreating acetylene and methyl acet-ylene/propadiene (MAPD), and sepa-rating of various components, such as ethylene, propylene, methane, propane and hydrogen. There are three types of olefin plants as defined by the recovery sequence, first separation step and po-sition of the acetylene hydrogenation; these process methods include the:

• Front-end demethanizer (DeC1) system

• Front-end deethanizer (DeC2) system

• Front-end depropanizer (DeC3) system.

The front-end DeC1 system is the most commonly adopted scheme. The

process is used predominantly by earlier-constructed ethylene plants. It can be applied to all feedstocks from ethane to gasoil (GO). Other processing schemes are preferred when considering high ef-ficiency and when feedstock flexibility to process heavier feeds such as propane, naphtha and GO are a priority.l In the presented study, the DeC3 system will be evaluated; the new system applies a two-tower subsystem positioned immediately after the CGC and drying steps.

FIG. 1 shows the integrated CGC and front-end DeC3 system of an ethylene plant. After the naphtha feedstock is cracked, it is sent to the oil-quench and water-quench towers. In the quench sys-

tem, the cracked gas is cooled and partially condensed. The quench-tower overhead vapors are sent to a four-stage CGC sec-tion for compression and drying before being directed to the DeC3 subsystem.

The DeC3 subsystem is used to sepa-rate C3 and lighter components from the charge gas. It consists of two distillation columns: a HP depropanizer (HP DeC3) operated at 12 bar, and the low-pressure depropanizer (LP DeC3) operated at 7 bar. The C3 and lighter components are removed from the top of HP DeC3, while C4 and heavier components exit from the bottom of the LP DeC3.

By applying the HP and LP DeC3 sub-systems, the olefin-unit operational flex-

EV-01

CGC 1st CGC 2nd CGC 3rd

EV-12EV-02 EV-04

EH-01 EH-02 EH-03EH-11 EV-06

ER-50

ER-60

ET-10

EV-42

CGC 4th

ER-32

EH-52

EH-51ER-05

EH-53

ER-31EH-56EH-58

EV-41

ET-11

EH-61

EV-40

EP-20

Feedfromquench

Minimum flow bypass

Anti-surge recycle

Vapor to DeC1

EV-16

ET-16

To waterquench

To water quench

To DeC4

Liquid to DeC1

Anti-surgerecycle

To DeC4

ET-12

EV-03

EP-16

EH-16

EH-59

EH-60

EH-50

FIG. 1. Flow diagram of the integrated CGC and front-end DeC3 system.


Petrochemical Developments

ibility can be improved when heavy feed-stock, such as naphtha, is used. The system is considered cost-effective because most of the C3 and lighter components can be removed from the feed stream and sent to the LP DeC3 tower. This design signifi-cantly reduces the refrigeration duty and minimizes fouling problems, which are experienced in a one-tower system.

In addition, there is no need for exter-nal hydrogen supply for the acetylene re-actor, since the overhead stream from the HP DeC3 contains enough hydrogen for hydrogenation. Note: The fourth stage of the CGC is actually located after the over-head drum of the HP DeC3. It only com-presses the C3 and lighter components; this also reduces the compressor loading as compared to a back-end DeC1 process, which compresses everything before the cracked gas enters the demethanizer.

Process description. As shown in FIG. 1, the CGC system consists of four stages with intercooling, liquid separa-tion, pumps and various auxiliaries. The cracked gas from the quench is fed to the first-stage compressor through a suction drum, and its pressure is raised from 1.28 bar to 3.36 bar. An interstage cooler chills the compressed cracked gas to 20°C. Similarly, each of the compression stages

consists of a suction drum, i.e., EV-01, EV-02, EV-12 and EV-42, a compres-sor (first stage to fourth stage), and an aftercooler, i.e., EH-01, EH-02, EH-03 and EH-51. The pressure of the cracked gas is raised to 7.76 bar after the second compression, 14.87 bar after the third, and 39.7 bar after the fourth. In the CGC flash drums, the cracked-gas mixture is separated into two phases: oil-liquid and vapor; or into three phases: water, oil-liquid and vapor. Condensates from the second-stage discharge drum (EV-03), the third-stage suction drum (EV-12) and the dryer knockout drum (EV-06) are sent back to the second-stage suction drum (EV-02), where the condensed liq-uid is fed to the CGC stripper (ET-16) and the vapor goes to the second-stage compressor. Condensate from the third-stage discharge drum (EV-04) is recycled to the third-stage suction drum (EV-12).

Other than the listed units, a caustic wash tower (ET-12), and a series of dry-ers and dehydrators are included in the CGC section. Between the second and third stages, the cracked gas is treated by a caustic tower to remove acid gases—carbon dioxide (CO2) and hydrogen sulfide (H2S)—to mitigate corrosion and freezing problems in the cold-separation section. The liquid hydrocarbon from the

second-stage suction drum (EV-02) is fed to the CGC stripper (ET-16), where the overhead is sent back to the water-quench tower, while the bottoms are pumped to the debutanizer tower. After the third-stage discharge drum (EV-04), a dryer feed cooler (EH-11) chills the cracked gas to 13.5°C. Next, the dryer knockout drum (EV-06) removes water, and the heavy hydrocarbons are sent back to the second-stage suction drum. The cracked-gas dryer (ER-50) and cracked-gas dehydrator (ER-60) further eliminate water from the cracked gas before the gas enters the HP DeC3 tower (ET-10).

The cracked gas from CGC is cooled to –14°C by the HP DeC3 feed chiller (EH-50) prior to feeding to tray 21 of HP DeC3. The overhead stream from HP DeC3 is sent through the fourth-stage compressor to reach a pressure of 39.7 bar. A guard bed (ER-32) removes arsenic from the cracked gas. A series of acetylene reactor feed/effluent heat exchangers (EH-51 and EH-52) are arranged to warm the feed to the reaction temperature; the acetylene reactor (ER-05) hydrogenates acetylene from the HP DeC3 overhead.

Effluent from the acetylene reactor is routed back through the acetylene-reac-tor effluent cooler (EH-53), and sent to a molecular-sieve-dryer guard bed (ER-31) to remove any traces of remaining wa-ter. The dryer effluent is further cooled in the HP DeC3 reflux chiller (EH-56), and then partially condensed by C3 refriger-ant to approximately –33.5°C in the HP DeC3 condenser (EH-58), before enter-ing the HP DeC3 reflux drum (EV-41). The condensed HP DeC3 overhead leaves the reflux drum and is sent to the top tray of the HP DeC3. The vapor overhead from the reflux drum, containing C3 and lighter components from the cracked gas, is sent to the DeC1 and cold-box section.

To avoid excessive bottom tempera-tures, some C3 components are allowed in the bottom of the HP DeC3. High bot-tom temperatures tend to induce fouling of the reboiler and tray due to polym-erization of C4 and C5 dienes. The HP DeC3 is reboiled by quench oil in the HP DeC3 reboiler (EH-59).

The bottoms from the HP DeC3 are sent to tray 24 of the LP DeC3 (ET-11), where the overhead is totally condensed to 3°C against a tertiary refrigerant in the LP DeC3 condenser (EH-61). The con-densed overhead is sent to the LP DeC3

EP-addCooling waterPropylene refrigerantLP pressure steamQuench oilQuench waterElectricityHP steamExtra pipeline

Water

EV-01

CGC 1st CGC 2nd CGC 3rd

EV-12EV-02 EV-04

EH-01 EH-02 EH-03EH-11 EV-06

ER-50

ER-60

EH-50ET-10

EV-42

CGC 4th

ER-32

EH-52

EH-51ER-05

EH-53

ER-31EH-56EH-58

EV-41

ET-11

EH-61

EV-40

EP-20

Feedfromquench

Minimum flow bypass

Anti-surgerecycle

Vapor to DeC1

EV-16ET-16

To waterquench

To water quench

To DeC4

Liquid to DeC1

Anti-surgerecycle

To DeC4

ET-12

EV-03

EP-16

EH-16

EH-59

EH-60

FIG. 2. New design of the integrated CGC and front-end DeC3 system.



reflux pumps (EP-20) and is split to pro-vide reflux for both the HP and LP DeC3s. The larger portion of the LP DeC3 reflux is sent to the top tray of the LP DeC3. The remainder is routed to tray 6 of the HP DeC3. The LP DeC3 is reboiled against LP steam in the LP DeC3 reboiler (EH-60). The outlet material from LP DeC3 bot-tom, containing the C4 and heavier com-ponents, is sent to the debutanizer section.

NOVEL DESIGNThis study focused on energy sav-

ings by improving the design of an inte-grated CGC-DeC3 system. To study the integrated system behavior, four stages of compressors with suction drums, caustic wash tower, cracked-gas dryer, HP and LP depropanizers with reboilers and condens-ers, C2 convertor and other related process units are included in the case study model.

New design configuration. The gen-eral idea of the new design is to introduce heavy streams from the early compression sections directly into the fractionation sec-tion. Results: Loading of most processing units can be reduced, thus achieving en-ergy savings for compression work, heat-ing/cooling duty and electricity. However, some processing requirements must be satisfied. First, the heavy stream must be acid-gas free before entering the fraction-ation sections. Second, the streams must be dried before entering the DeC3 column. Third, certain pressure increments for the heavy streams may be needed to make them viable as feed to the DeC3 columns.

FIG. 2 is one innovative design, where condensate streams from the third-stage suction drum EV-12 and discharge drum EV-04 are sent into the HP DeC3 directly. A caustic wash (ET-12) ensures no H2S or CO2 contamination. To carry out the new design, several major changes were considered:

• One pump (EP-ADD) is added to increase the pressure of the first heavy stream from 7.73 bar to 14 bar.

• One dryer (EV-ADD) is required to knock out water from the heavy stream.

• The feed tray of the heavy stream to the HP DeC3 must be identified and optimized.

• Operating conditions of both LP and HP DeC3 columns, e.g., reflux ratio, boil-up ratio, are optimized.

Based on this design, several improve-ments are expected:

• A significant amount of compress-ing work on heavy components in the second and third stages can be saved.

• Operation loading of the CGC stripper can be reduced due to a lower inlet flowrate, including a work reduction in the pump.

• The cooling duty for the CGC third-stage aftercooler can be reduced.

• Improvements in other units, in-cluding pumps and heat exchangers, are achievable.

Simulation and optimization results. The optimization work of the new design is conducted with commercially available simulation software.a By establishing a steady-state model of the integrated CGC-DeC3 system, the operation parameters are tested. Likewise, the separation specifi-cations and safety requirements are moni-tored before the optimal results are ob-tained. For example, in the CGC section, the inlet/outlet pressure and temperature of each compressor stage are carefully han-dled to maintain a defined range.

TABLE 1 summarizes the operating con-straints of the DeC3 subsystem. There are specifications for the top temperatures and pressures of the two columns. More importantly, both towers are designed to operate with a bottoms temperature less than 82°C. This low temperature re-

duces fouling in the trays and reboilers. The calculated results from this study are 74.62°C of the HP DeC3 bottom, and 78.54°C of the LP DeC3 bottom; both are within the specs.

The DeC3 subsystem is the first sepa-ration unit of the whole fractionation section. It removes C3 and lighter compo-nents from the heavier ones, which will cause critical impacts to downstream pro-cessing if not operated properly. The C4 fraction in the top of HP DeC3 is limited to less than 100 ml/m3, which in this study, is reduced to 10 ml/m3. Some C3s are al-lowed in the bottoms of both columns to maintain low temperatures, thus mitigat-ing fouling problems. The specification of C3s in the HP DeC3 bottom is less than 37 mol%, and the calculated value is 14.29 mol%. It requires less than 1,095 ml/m3 of C3s in the bottom of LP DeC3. Study re-sults indicate 914.61 ml/m3. Other than listed specifications, the C2 fraction in the bottom of HP DeC3 is limited to less than 10,951 ml/m3. The result from this study is 3,580.20 ml/m3, which meets process-ing requirements. In summary, the calcu-lated data are all qualified with respect to the specifications listed in TABLE 1.

TABLE 2 indicates the energy usage re-sults with the new and the old designs, along with possible energy reductions. According to different utilities, seven

TABLE 1. Operating constraints of HP DeC3 and LP DeC3 columns

Column Description Model data Comment

HP DeC3

(ET-10)

Top temperature, –30±3 °C –31.51 Qualifi ed

Bottom temperature, 71±5 °C 74.62 Qualifi ed

Pressure, 12±2 bar 12.2 Qualifi ed

C4 in top, < 100 ml/m3

1,3-Butadiene 3.52

10 Qualifi edIsobutylene 6.29

n-Butane 0.19

C3 in bottom, < 37 mol%

Propadiene 0.94%

14.29% Qualifi edPropylene 13.07%

Propane 0.28%

C2 in bottom, < 10,951 ml/m3

Acetylene 166.08

3,580.2 Qualifi edEthylene 2,424.31

Ethane 989.81

LP DeC3

(ET-11)

Top temperature, 7±5 °C 2.9 Qualifi ed

Bottom temperature, 75±6 °C 78.54 Qualifi ed

Pressure, 7±0.8 bar 6.5 Qualifi ed

C3 in bottoms, < 1,095 ml/m3

Propadiene 909.36

914.61 Qualifi edPropylene 3.46

Propane 1.79



different colors are used to represent cooling water, C3 refrigerant, LP steam, quench oil, quench water, HP steam and electricity. Detail locations of each utility usage in the integrated CGC and DeC3 system are indicated in FIG. 2.

Results in the “Saving %” column of TABLE 2 show that almost half of the ener-gy consumption units can improve perfor-mance via reduced energy requirements. For example, the heating duty of the CGC stripper reboiler can be reduced by 42.1%, and the work required by the CGC strip-per pump can be reduced by 31.3%. The cooling duty of the CGC second- and third-stage aftercoolers can have a reduc-tion of 9.6% and 22.7%, respectively.

In the CGC, the most significant im-provements are achieved by the second- and third-stage compressors, which can reduce needed work by 0.4 MMBtu/hr and 0.5 MMBtu/hr. There is a negligible increment of 0.1 MMBtu/hr on the cool-ing water duty, a small amount of addi-tional heating duty, 1.2 MMBtu/hr from quench water, and a tiny increment of 0.002 MMBtu/hr on electricity.

The energy savings are significant with the new design as compared with

previous designs. For example, the total amount of saved cooling duty from C3 refrigerant is 4.1 MMBtu/hr—a 10% re-duction. This suggests significant reduc-tions in the workload for the C3 refriger-ant system. Meanwhile, the heating duty required for LP steam can be reduced by 2.8 MMBtu/hr, which is a 26.9% savings as compared to the previous design. In ad-dition, a 2.4 MMBtu/hr reduction saves 75.3% heating duty for the CGC stripper reboiler under the new design. Finally, the amount of work needed for the compres-sors can be reduced by 0.87 MMBtu/hr in total, which saves 2.1% from the previous consumption of CGC work.

ACKNOWLEDGMENTSThis research work was supported in part by the

Texas Air Research Center, Texas Hazardous Waste Research Center and the Graduate Student Scholarship from Lamar University.

NOTES a The optimization work of the new design is carried

out with Aspen Plus 7.3, Aspen Technology, Inc., 2011.

LITERATURE CITED 1 Falqi, F., The Miracle of Petrochemicals—Olefins

Industry: an In-Depth Look at Steam-Crackers, Universal Publishers, Boca Raton, Florida, 2009.

JIE FU is a PhD candidate with the Dan F. Smith Department of Chemical Engineering at Lamar University. He holds a BS degree from the East China University of Science and Technology and was an exchange student at the University

of Houston. His research interests include process design, dynamic simulation and production scheduling for energy saving and emission reduction in industries.

CHUANYU ZHAO is a PhD candidate with the Dan F. Smith Department of Chemical Engineering at Lamar University. She received her BS degree from the East China University of Science and Technology, and was an exchange

student at the University of Houston. Her research areas include process modeling, planning, scheduling and optimization, process simulation, process safety and process synthesis with particular applications in petrochemical, refining and electroplating industries.

QIANG XU is an Associate Professor of the Dan F. Smith Department of Chemical Engineering and 2012 University Scholar at Lamar University. He holds BS and MS degrees along with a PhD from Tsin-ghua University, China, all in chemical

engineering. His research involves process modeling, scheduling, dynamic simulation and optimization, industrial pollution prevention and waste minimization, and chemical process safety and flexibility analysis. His research work has been extensively supported by indus-tries and by federal and State of Texas funding agencies.

TABLE 2. Energy consumption optimization results summary

Unit Description Utility, MMBtu/hr Previous design New design Saving Saving, % Total savings

EH–53 C2 Reactor effl uent cooler Cooling water –9.73 –9.76 –0.03 –0.3

–0.1 (0.2%)

EH–01 CGC 1st-stage aftercooler Cooling water –23.63 –26.43 –2.8 –11.8

EH–02 CGC 2nd-stage aftercooler Cooling water –9.93 –8.98 0.95 9.6

EH–03 CGC 3rd-stage aftercooler Cooling water –18.08 –13.98 4.1 22.7

EH–11 Dryer feed chiller Cooling water –0.63 –0.79 –0.18 –25.6

EV–12 CGC 3rd-stage suction drum Cooling water –0.33 –2.53 –2.2 –668.1

EH–50 HP depropanizer feed chiller C3 refrigerant –8.54 –6.85 1.69 19.8

4.1 (10.0%)EH–58 HP depropanizer condenser C3 refrigerant –13.51 –13.54 –0.03 –0.2

EH–56 HP depropanizer refl ux chiller C3 refrigerant –12.35 –12.44 –0.09 –0.8

EH–61 LP depropanizer condenser C3 refrigerant –6.56 –6.48 0.08 1.3

EH–52 Acetylene reactor feed heater LP steam 4.49 4.50 –0.01 –0.22.8 (26.9%)

EH–60 LP depropanizer reboiler LP steam 6 5.88 0.12 2.0

EH–59 HP depropanizer reboiler Quench water 5.72 6.80 –1.08 –18.9 –1.2 (21.5%)

EH–16 CGC stripper reboiler Quench oil 3.22 1.87 1.36 42.1 2.4 (75.3%)

CGC 1st CGC stage 1 HP steam 11.94 11.94 0.0 0.0

0.87 (2.1%)CGC 2nd CGC stage 2 HP steam 10.37 9.97 0.40 3.9

CGC 3rd CGC stage 3 HP steam 7.59 7.09 0.50 6.6

CGC 4th CGC stage 4 HP steam 11.28 11.31 –0.03 –0.3

EP–16 CGC stripper pump Electricity 0.024 0.016 0.01 33.3

–0.002 (4.7%)EP–20 LP depropanizer refl ux pump Electricity 0.019 0.018 0.001 5.3

EP–Add Added pump Electricity 0.00 0.011 –0.01 –

T H E E X P E C T E D .

We provide solutions. Smith & Burgess is the expert in flare and relief system design. Our methodology is cost-effective and we pride ourselves on using realistic assumptions that meet industry standards, yet don’t burden the owners with excessive concerns. We are committed to our customers. At Smith & Burgess we are passionate about protecting your assets. We insist on quality. If that’s beyond what you expected, we did our job right.

w w w . s m i t h b u r g e s s . c o m


BETE...

your one-stop resource

for drop-in spray lances

and quills.

BETE. Your strategic partner for engineered spraying solutions.

CFD Analysis

Distribution

Evaporation

Performance Through Engineering BETE is your complete single-source supplier of custom spray lances, quills, and chemical injectors. Our in-house operation includes all aspects of design, fabrication, and performance testing, ensuring a seamless process from RFQ to delivery.

BETE’s welders have decades of extensive experience in welding dissimilar metals and exotic alloys and are fully qualified to Section IX. BETE’s integrated engineering, quality, and manufacturing departments combine to meet a wide variety of code, testing, and inspection requirements. ASME B31.3 fabrications are our specialty.

Request a lance quote...BETE’s Application Engineers are ready to take your sketch/inquiry and produce a recommendation and drawing of our proposed solution. We’ll use our experience to work with you to improve your process with the right spray nozzle coupled with custom fabrications designed for your application.

MaxiPass provides a full cone spray pattern

with even distribution

Made in the USA

BETE Fog Nozzle, Inc.T +1(413)772-0846F +1(413)772-6729www.bete.com

MaxiPassTM NozzlesThe ultimate in clog-resistance with the largest free passage available in a full cone nozzle

SpiralAirTM NozzlesA unique three-stage process that uses atomizing air more efficiently than conventional designs, resulting in better atomization at high liquid flow rates

Spray LancesCustomized spray nozzle lance assemblies are used for scrubbing, quenching, injection and washing that meet the strictest quality and performance criteria

Retractable Spray Lance



Bonus Report Petrochemical Developments J. SHIN, J. LEE and S. LEE, LG Chem Ltd., Daejon, South

Korea; B. LEE, AMT Pacific Co. Ltd., Seoul, South Korea;

and M. LEE, Yeungnam University, Gyongsan, South Korea

Enhance operation and reliability of dividing-wall columns

Dividing wall columns (DWCs) are an attractive option when the goals are to reduce energy consumption and capital investment for distillation processes. A major component in a DWC system is the reflux splitter, which separates the liquid products from the column overhead into two liquid streams: one to the prefractionation section and the other to the main column section of the DWC. A more reliable reflux-splitting system stabilizes the entire column, thus minimizing unit/pro-cess downtime. Several novel developments for a reflux-split-ting system can improve the operation and reliability of DWCs.

Background. DWC technology has been successfully applied in various petrochemical processes.1,2 A DWC can reduce en-ergy consumption by removing the remixing phenomenon within a conventional two-column system for a ternary mix-ture. DWC systems can minimize capital investment by inten-sifying a conventional two-column system into a single column with a simple dividing wall.

The concept of DWC was proposed in 1949; however, it took a long time to adopt this principle.3 Availability of proper design

tools is a major obstacle in developing more complex separation-column systems as compared to a conventional column. Since the 1980s, powerful CAD software enables the design and com-mercialization of DWCs for a range of petrochemical processes.

Once the structure of a DWC is fixed for a given product specification, the most important optimizing variables remain-ing are the liquid and vapor split in the dividing wall section, as shown in FIG. 1. These variables have a significant effect on the total separation performance in the DWC.4 FIG. 2 shows the effect of internal flow distribution on the energy consump-tion in a typical DWC. This figure illustrates the existence of an optimal internal flow distribution that generates the lowest energy consumption. Generally, the internal liquid and vapor flows into the prefractionator and main dividing wall section are the most crucial design factors. Both impact the total en-ergy consumption and separation efficiency. The energy effi-ciency of DWC can drastically deteriorate by a small deviation in the internal flows from the optimal conditions. Conversely, only the liquid split can be adjusted during operation. The va-por split cannot be manipulated arbitrarily once the column is

0.0

0.6

1.2

1.8

2.4

3.0

3.6

4.2

4.8

5.4

6.0

1015

2025

3035

40

2530

3540

4550

55

Rebo

iler h

eat d

uty,

MW

Preliquid flowrate, kgmole/hrPrevapor internal flowrate, kgmole/hr

FIG. 2. Typical effects of liquid and vapor split on energy efficiency in DWCs.

Main column

Dividing wallPrefractionation

section

Vapor split

Liquid split

FIG. 1. Schematic diagram of a typical DWC.



constructed. For this reason, in most commercial DWCs, there should be a device to split the liquid coming from the column overhead into sections divided by a wall, i.e., a prefractionation section and main column section.

Several reflux splitters have been adopted in commercial DWC installations. A simpler and more reliable reflux-splitting system is the optimum choice for better operability and reliabil-ity of the total column system. A novel reflux-splitting system can provide better operation and a more energy-efficient DWC.

Internal reflux splitters for DWC. Conceptually, the split-ting of a liquid stream can be achieved by a simple arrangement of a pump, valves, pipes and vessel, as shown in FIG. 3. Nev-ertheless, it still requires additional plot area, equipment and instruments. The increased equipment inventory within the system should be minimized. For this reason, this concept is barely used in practical applications of DWCs.

The ideal reflux-splitting device for DWC applications should have these technical features and requirements, consid-ering its purpose and mounting location:

• Design does not include a pump and a moving/rotating unit.• Height and volume of the column should be minimized.• System can accurately manipulate and control the split ratio.• Unit is simple, durable and reliable.

Moving bucket reflux splitter. First-generation reflux-split-ting systems are shown in FIG. 4.5,6 These early systems consist of a casing that is subdivided into three chambers. The feed chamber is located above where the liquid is directed onto either the prefractionation or main column chamber depend-ing on the positioning of the dividing body. The split ratio is controlled by adjusting the timing of the hollow inner bucket into the prefractionation or the main column chamber. The dividing body is actuated by magnetic coupling, which allows a pressure- and vacuum-tight design. The exterior drive is a pneumatically driven rotary motor.

Manual reflux splitter. The moving elements in the first-gen-eration reflux splitter can be damaged by the constant movement of the mechanical parts. In a petrochemical plant, the splitter should operate at least 8,400 hr/yr without mechanical prob-lems. For example, the naphtha cracking center (NCC) has more severe requirements to address possible equipment malfunction because it needs to have run length of at least three to four years without regular scheduled plant maintenance. A more reliable liquid-splitting device should be used in such cases. For this purpose, a manual-type reflux splitter that requires no moving element was devised, as shown in FIG. 5, and it was successfully installed on several commercial DWCs in LG Chem plant sites.7

The internal structure of the manual reflux splitter appears complex, but the basic principle is quite simple. As shown in FIG. 5, the splitter is divided into several sections that receive the liquid drawoff from the column overhead. Note: Each sec-tion has a liquid distributor with a different number of holes to adjust the liquid flowrate.

An on/off type of automatic valve is mounted to each sec-tion. If the valve is closed, the received liquid will overflow into the main section, which is connected directly to the main column without a valve. By combining the opening of each valve, the split ratio can be controlled by predetermined val-ues. The control range of the split ratio can be determined

FIG. 4. Reflux splitters with moving elements.5, 6 FIG. 5. Cutaway view of a manual reflux splitter.

FC

Vessel

FIG. 3. Conceptual design of liquid splitting in a DWC.



from a sensitivity study of the DWC column for possible op-eration scenarios, such as variations in feed compositions and product quality. The performance of the reflux splitter can be verified before mounting it into the DWC column using a water-run test.

In-the-column reflux splitter. The latest develop-ment in the reflux-splitter system increases unit reli-ability with a simple modification of the conventional distributor and collector tray, as shown in FIG. 6.8 The in-the-column reflux splitter consists of a modified collector tray, collector box and two transfer lines. The collector tray gathers the liquid from the top sec-tion and transfers it to the collector box located below.

Holes in the middle of the collector tray (see FIG. 6B) divide the liquid into four sections in the collector box with predefined fractions. As shown in FIG. 6C, the splitting ratio can be adjusted by the on/off operation of two valves, which are located on the transfer lines (see FIGS. 6A and 6C) at the out-side of the column. If the two valves are open, then the liquid stream from the middle sections goes to the pre-section via liquid transfer lines, which are connected to the distributor for the presection. Thus, it will increase the liquid flow to the pre-section. If the two valves are closed, the liquid flow to the main section will be increased by the liquid overflow to the main sec-tion. Four combinations in the on/off valve are possible, and, thus, the four split ratios are available. In the design phase, the required split ratios can be predefined considering possible op-erational scenarios.

The developed reflux splitter satisfies all the technical re-quirements or specifications described earlier; it can reduce fabrication and installation costs. This splitter type is simpler and cheaper than the manual reflux splitter, as well as the buck-et splitters. In addition, it can reduce the column height, thus providing the opportunity to install more trays or packing sec-tions with a longer length.

The operating range of this splitter can be predetermined and the performance can be verified before installation in the col-

umn. As shown in FIG. 7, the reflux-split ratio can be controlled accurately by closing and/or opening the valves. LG Chem’s ex-perience from several petrochemical applications suggests that,

in most cases, only four combinations are sufficient to control the split ratio for expected operational variations.

Evaluation. Several types of internal reflux splitters were ana-lyzed in terms of reliability, structural simplicity and ease of maintenance. TABLE 1 summarizes the results of the study com-paring three different reflux splitters.

Recent developments in the internal reflux-splitting sys-tem make the DWC application more reliable with a simpler structure. The reflux splitter without a moving element may not cover a wider range of split ratios, but it can provide more robust and reliable operation with less maintenance. Overall,

FIG. 6. In-the-column reflux splitter: A) overall, B) collector tray and C) collector box and transfer lines.

DesignActual, min. flowActual designActual, max. flow

1.51.5 2.0 2.5 3.0 3.5

2.0

2.5

3.0

3.5

FIG. 7. Results of the water-run test for the in-the-column reflux splitter.

A new internal reflux-splitter system with

no moving part can achieve more reliable

operation of DWCs. A DWC can reduce

energy consumption by eliminating the

‘remixing’ phenomenon with a conventional

two-column system into a single column.

TABLE 1. Comparisons of several refl ux-splitter systems

Splitter type Operating window/control

Reliability/maintenance

Cost/structure

Moving bucket Wide/continuous Fair/overhaul needed

High/complex

Manual Narrow/discrete Excellent/regular valve inspection

Moderate/simple

In-the-column Narrow/discrete Excellent/regular valve inspection

Low/simple



system specific design with a modeling study for possible op-erational scenarios is recommended to overcome the flexibility issues of the newly developed splitter.

ACKNOWLEDGMENTThis study was supported by the Basic Science Research Program through

the National Research Foundation of Korea (NRF) funded by the Ministry of Education, Science and Technology (2012012532).

LITERATURE CITED 1 Asprion, N. and G. Kaibel, “Dividing wall columns: Fundamentals and recent

advances,” Chemical Engineering Progress, Vol. 49, 2010, pp. 139–146. 2 Shin, J., S. Lee, J. Lee and M. Lee, “Manage risks with dividing-wall column instal-

lation,” Hydrocarbon Processing, June 2011, pp. 59–62. 3 Wright. R. O., “Fractionation Apparatus,” US Patent 2,471,134, May 1949. 4 Lee, S., M. Shamsuzzoha, M. Han, Y. Kim and M. Lee, “Study of structure char-

acteristics of a divided wall column using the sloppy distillation arrangement,” Korean Journal of Chemical Engineering, Vol. 28, 2011, pp. 348–356.

5 http://www.montz.de/sites/products/reflux.fr.html. 6 http://www.nascentprocess.net. 7 Lee, B., G. Kim, M. Lee, J. Lee, J. Shin, and S. Lee, “Liquid splitter,” Korea patent

pending, 10-2010-0120467, November 2010. 8 Lee, B. and G. Kim, “Liquid splitter,” Korea patent pending, 10-2012-0030223,

March 2012.

JOONHO SHIN is a process systems engineer at LG Chem Ltd., in South Korea. In 1997, he began his professional career as a design and control specialist with SK Engineering & Construction. Dr. Shin’s industrial experience has focused on modeling, optimization and control of chemical and petrochemical industrial plants. He holds a BS degree in chemical engineering from Korea University, and an MS degree and PhD in chemical

engineering from the Korea Advanced Institute of Science and Technology (KAIST).

SUNGKYU LEE is a process systems engineer at LG Chem Ltd. in South Korea. He holds BS and MS degrees in chemical engineering from Chungnam National University. He has worked on modeling, optimization and control of chemical and petrochemical plants since 2002.

JONGKU LEE is the vice president of LG Chem Ltd. at Research Park, South Korea. He holds a BS degree in chemical engineering from Seoul National University, and an MS degree and PhD in chemical engineering from KAIST. Since joining LG Chem in 1994, he has worked on numerous process modeling and optimization projects. Dr. Lee is in charge of LG Chem’s process modeling and solutions group and is performing

research in the areas of energy saving and sustainability in chemical plants.

BYEONGKYEOM LEE is a chief technology officer of process engineering, mass-transfer equipment design and application for AMT Pacific Co. Ltd., in South Korea, since 2000. He holds a BS degree in chemical engineering from Sungkyunkwan University. He began his professional career as a process engineer at Kolon Engineering Inc. in 1989, and worked in SK Engineering & Construction Co. Ltd. from 1991 to 2000 as

a specialist of mass-transfer equipment design and application. He has over 20 years of experience in distillation column design and applications for refinery, petrochemical and chemical plants.

MOONYONG LEE is a professor at the school of chemical engineering at Yeungnam University in South Korea (http://psdc.yu.ac.Kr). He holds a BS degree in chemical engineering from Seoul National University, and an MS degree and PhD in chemical engineering from KAIST. Dr. Lee worked in SK Energy’s refinery and petrochemical plants for 10 years as a design and control specialist. His current areas of

specialization include modeling, design and control of chemical processes.


At Avondale, our highly skilled workforce is energized and set to deliver innovative solutions for upstream, midstream and downstream energy infrastructure projects. Backed by our vast facilities, unmatched capabilities, and 75-year legacy of success, we’re harnessing our energy to build solutions today that will power the world for years to come.

World-Class Engineering and Modular Constructionfor Energy Infrastructure Projects

Visit us at www.hii-avondale.com

504-654-5000 | 5100 River Road Avondale, LA 70094

THEY ALL FLOW THROUGH

AVONDALE

UPSTREAM,MIDSTREAM

ORDOWNSTREAM,


TERMINALS AND STORAGE

Overflow systems are the last line of defense T–92Terminals and storage news T–94

CORPORATE PROFILESInduMar Products, Inc. T–95 CB&I T–97

2013

Special Supplement to

T–92 TERMINALS AND STORAGE 2013 | HydrocarbonProcessing.com


OVERFLOW SYSTEMS ARE THE LAST LINE OF DEFENSEM. TOGHRAEI, Engrowth Training, Calgary, Alberta, Canada

The overflow system is the last line of defense against the excessive container filling. An overflow system consists of an overflow nozzle with internal and/or external piping and, in some cases, a siphon breaker. In this context, the container could be an atmospheric tank or a non-flooded vessel. Howev-er, not all the containers need an overflow system. If siphoning the liquid out of the container is not affordable, or if the other protective layers against overfilling provide sufficient coverage, an overflow system may not be necessary. In general practice, atmospheric tanks—but not vessels—have overflow systems.

Overflow nozzle location. Orienting the overflow nozzle tends to not be significant from a process perspective, unless the tanks possess complexities such as compartments. The overflow nozzle orientation can be decided based on plot plan preference (FIG. 1).

In a very basic design, the overflow nozzle may be located high on the side of the tank, usually above high high liquid level (HHLL) and below the roof seam. Subsequent research re-vealed that this design causes liquid splashing during the tank overflow, which is hazardous if the liquid in question is an “ag-gressive” liquid, meaning it is flammable, irritable or corrosive. To fix this problem, the overflow nozzle can be piped down to the bottom of the tank, where an elbow can prevent the tank foundation from washing out due to the overflow liquid (FIG. 2).

If a tank is blanketed at the top, such an overflow arrangement could provide an opportunity for the blanket gas to escape from the tank. To prevent this, the overflow nozzle could be extended internally to the lower part of the tank. One option is to lead this

pipe down to below the low low liquid level (LLLL) in order to ensure the presence of a liquid at the inlet of the overflow, preventing the escape of the blanket gas. Care should be taken to not extend the internal pipe to the very bottom of the tank, especially if the stored liquid is not very clean and sludge may be present at the bottom of the tank, which could clog this pipe.

Additionally, there may be cases where it is required to have both an internal pipe and an external one, in which case it is important to utilize a siphon breaker. In the absence of such a breaker, the overflow stream will continue even after the liquid has dropped to below the threshold for the overflow nozzle. In other words, the tank will continue to overflow even after recovering from an upset.

Overflow sizing. Proper overflow nozzle sizing relies on a careful definition of its purpose within a refinery’s needs. The nozzle should provide enough area to discharge the flow out of a container to protect it from overflowing, when all other outlets to the container are blocked and all the inlets are pass-ing. Alternatively, some companies define an overflow scenario as one where all the nozzles are blocked except for the biggest inlet nozzle. The second definition is obviously less conserva-tive than the first one. Therefore, the first step of overflow sizing is deciding on the flow (or summation of flows) that causes a container to overflow.

To develop a methodology for overflow sizing, first observe the phenomenon. When all the outlet nozzles are blocked, the liquid level in the container starts to rise from normal liquid lev-el (NLL). This continues until the liquid level reaches the bot-tom of the overflow nozzle, at which point the tank begins to overflow. Because the overflow nozzle is not sealed by flow, the overflow area is sufficient and the tank level continues to rise.

Assume for the sake of discussion that the overflow nozzle area is not adequate for overflow and that the tank level con-tinues to rise. This is significant if and when the liquid level reaches the seam of the roof. At this point, the flow out of the overflow nozzle should be equal to the flow that comes into the tank. In such a case, the overflow nozzle fails if and when it is unable to accommodate a flow equal to inlet flow and the liquid level inside of the tank reaches the roof space.

Assume a hypothetical draining container exists at the top of the tank. The solution is:

Q Cd A 2 2 gh (1)

Q is the inlet governing flowrate , Cd is the discharge co-effi-cient (which can be considered as 0.6 for a sharp-edged outlet) and h is the distance between line from the overflow nozzle to the top of the tank. From Eq. 1, the overflow area (A) and the overflow inside can be calculated (FIG. 3).

Elevation

Side viewPlan view

Orientation

FIG. 1. Plot plan preference

FIG. 2. The overflow nozzle can be piped down to the tank bottom.

HYDROCARBON PROCESSING | TERMINALS AND STORAGE 2013 T–93


Example 1. Determine the overflow size for a water tank with 15 m height and 24 m diameter. The governing in-flow is 850 m3/hr as shown by:

Q Cd A 2 2 gh (2)

Assume that the height of liquid above the centerline of the overflow nozzle is 0.5 m, then Eq. 3 is:

(3)850

3,6000.6 A 2 2 9.8 0.5

D = 15.8 in. or 16 in.When a pipe is connected to the overflow nozzle, the same

formula can be used with a slight correction:

(4)

Q =Cd

1+ f LD

×A× 2 gh

f is the moody friction factor and L is the total actual and equivalent length of overflow pipe with D as the diameter.

Two-phase flow. Another aspect of overflow nozzle needs to be addressed —two-phase flow. Initially, the overflow system (nozzle and downcomer) is full of air until the onset of over-flowing when a partial flow will appear. If the level goes higher and flow increases, the liquid stream and air will generate a two-phase flow. There is not much research on vertical downflow liquid-gas two-phase flows; however, three regimes have been identified in such an arrangement.1

At a low liquid rate, the regime will be “annular flow” for a short period, “slug flow” at a moderate rate and “bubble flow” at a high liquid rate (FIG. 4).

Because the air pressure is not high, there is a chance that in the last stage of the regime, it will be converted into a single, liquid phase. This actually is preferable, because if the regime is annular, then the capacity of the overflow system is reduced; and in the two other regimes there would be a big vibration in the overflow pipe. This favored regime conversion can happen if the liquid velocity is higher than the “sweeping” velocity. To ensure the liquid velocity is higher than sweeping velocity, it is recommended that Eq. 5 is followed:2

Fr ≥ 0.31 (5)

Fr is the liquid-phase Froude number: Fr =

VL

g×D2×

ρL

ρL−ρG (6)

However, air density is negligible in comparison to liquid density and the Froude number can be reduced:

(7)

Fr =

VL

g×D2

Therefore, the Froude number of overflow should be greater

than 0.31. Otherwise, the overflow nozzle should be decreased, and the liquid head above nozzle should be increased accordingly.

Example 2. In Example 1, make sure there is minimum vibra-tion during overflowing:

(8)

8503,600

π4× 16×

25.41,000⎛

⎝⎜⎜⎜

⎞

⎠⎟⎟⎟⎟

⎡

⎣⎢⎢

⎤

⎦⎥⎥

2 =1.82 m/sec

1.8

9.8 16× 25.41,000

⎡

⎣⎢⎢

⎤

⎦⎥⎥

2 = 0.91>0.31

With Eq. 8, the air slugs should be swept away. Another

method to ensure there is no entrainment in the overflow downcomer is:3

Fr 1.6 hD

2

(9)

Siphon breaker sizing. A siphon breaker is a pipe that is located at the highest point of the inverted U and is routed to the atmo-sphere, outside or inside of the tank. This pipe allows the system to suck the gas from outside the system. This breaks the siphon and prevents flow out of the overflow nozzle, due to the siphon phenomenon. As a rule of thumb, this pipe is a straight pipe with a length of around 4 m to 5 m and a diameter of 2 in. to 3 in. However, for the detailed design, it should be considered that a quick gas stream from the outside needs to “break the vacuum” by travelling through this pipe, so that there is an increase in the vacuum pressure (almost 0 kpag) to a pressure equal to the liquid column in the inverted U leg.

h

FIG. 3. Overflow nozzles can fail when the liquid level reaches the roof space.

Increasing liquid flow

Slug flowAnnular flow Bubble flow

FIG. 4. Three liquid rate possibilities.

T–94 TERMINALS AND STORAGE 2013 | HydrocarbonProcessing.com


This flow is in an adiabatic condition because of its rapid occurrence. In a simple system, the flow through a 2-in. pipe is about 0.29 kg/sec and is meant to fill out a specific space in the bend of the inverted U to break the vacuum. This space is theo-retically only a thin layer of gas in the inverted U cross-section which provides discontinuity in liquid siphon flow. However, in practical situations, it is possible that a larger space will be needed to fill out and discontinue siphon flow. A realistic approach is to assume that the destination space could equal the bend volume.

Ideal Gas Law PV = nRT can predict the required mass in this flow and these two numbers together provide the time to break the vacuum. This is the time that must elapse to stop the flow of siphon out of the overflow nozzle. Now it is up to the designer to decide if this time is acceptable or if the siphon breaker pipe should be enlarged.

Example 3. In Example 1, calculate the siphon breaker size. The source data is atmospheric pressure; assume a pressure of 101 kpag, an ambient temperature of 20°C, and the destination point data is pressure zero. The pipe is assumed to be 4 m long with a 2-in diameter.

The solution is based on calculating adiabatic compress-ible flow.4 N represents all the losses in pipe and is K + f (L/D). K is considered 1 for an exit and for friction factor ( f ), a fully turbulent flow friction factor is used (TABLE 1).5 For N equal to 2.5, flux will be 143.6 kg/s.m2 or 0.29 kg/s.

This flow should fill the bend volume. If the stretched length of bend is 5 m, its volume would be 0.65 m3. If the height of the inverted siphon leg is 10 m, the pressure in the bend will be:

P = ρgh’ = 1,000×9.8×101,000

= 100 KP absolute pressure

(10)

PV mM

RT

100 .065m29

0.008314 20 273

m = 1,554 gr

R 0.008314 m3kPa

mol. K

(11)

M = 29 for airT = 20°C = 293KGives m = 1,554 gr or 1.554 kgElapsed time would be:

(12) t =

mm°

= 1.5540.29

= 5.3 sec

During this 5.3 sec, 1.26 m3 of water will be wasted due to siphoning:

Volume = 5.3 � 850/3600 = 1.26 m3 or 0.02% of total tank volume (13)

LITERATURE CITED 1 Barnea, B. “Flow Pattern Transition for Vertical Downward Two Phase Flow”,

Chemical Engineering Science, Vol. 37, Issue 5, 1982. 2 Coker, A. K., FORTRAN: Programs for Chemical Process Design, Analysis and

Simulation, Gulf Professional Publishing, p. 182, 1995. 3 Perry, R. H., and D. W. Green, Eds., Perry’s Chemical Engineers’ Handbook, 6th Ed.,

McGraw-Hill, New York , pp. 6-28, 1984. 4 Perry, R. H., and D. W. Green, Eds., Perry’s Chemical Engineers’ Handbook, 6th Ed.,

McGraw-Hill, New York, pp. 6-23, 1984. 5 “Crane Technical Paper No. 410,” Crane Valves Co., New York, 1988.

MOHAMMAD TOGHRAEI is a consultant with Engrowth Training. He has over 20 years of experience in the field of industrial water treatment. His main expertise is in the treatment of wastewater from oil and petrochemical complexes.

TERMINALS AND STORAGE NEWSMETHANOL STORAGE IN LOUISIANA

Kinder Morgan Energy Partners has entered into a long-term contract with Methanex to support the construction of metha-nol storage capacity near Kinder Morgan’s Geismar liquids ter-minal (GLT) in Geismar, Louisiana. Kinder Morgan will build, own and operate the storage tanks and related infrastructure, including improvements to its existing dock at GLT. The assets will provide critical marine, rail and truck access in support of the 1 million tpy methanol production plant being relocated by Methanex from Chile.

The terminal infrastructure is expected to be in service dur-ing the second half of 2014, coinciding with the anticipated startup of the relocated plant. In a separate deal, Kinder Mor-gan also acquired Quality Carriers’ 26-acre terminal located in Chester, South Carolina. The 19-tank facility currently provides storage for a single customer, with a capacity of 35,000 bbl. The terminal receives product by rail and distributes by truck.

TESTIMONY BLASTS DHS EFFORTS AT ASSESSING CHEMICAL SECURITY RISK

The US Government Accountability Office (GAO) has found that the tiering approach used to regulate high-risk chem-ical facilities by the Department of Homeland Security (DHS) Infrastructure Security Compliance Division fails to properly consider each of the risk elements involved in a potential ter-rorist attack (threat, vulnerability and consequence). This con-clusion was presented on March 14 by GAO Director Stephen Caldwell during testimony before the US House Energy and Commerce Committee. A similar finding was also the basis for the International Liquid Terminals Association petition to DHS to remove gasoline facilities from regulation under the Chemi-cal Facility Anti-Terrorism Standards (CFATS). GAO estimates that it could take seven to nine more years for ISCD to complete its first review of security plans from all CFATS-regulated sites, which began during 2011.

TABLE 1. Pipe friction data for clean commercial steel pipe with fl ow in complete turbulence

Nominal size 1⁄2 in. 3⁄4 in. 1 in. 11⁄4 in. 11⁄2 in. 2 in. 21⁄2 in., 3 in. 4 in. 5 in. 6 in. 8–10 in. 12–16 in. 18–24 in.

Friction factor, fr 0.027 0.025 0.023 0.022 0.021 0.019 0.018 0.017 0.016 0.015 0.014 0.013 0.012

SPONSORED CONTENT HYDROCARBON PROCESSING | TERMINALS AND STORAGE 2013 T–95

INDUMAR PRODUCTS, INC.

IF IT LEAKS OR CORRODESINDUMAR CAN STOP IT

Since 1986 InduMar Products, Inc. has provided a unique solution with the STOP ITR PIPE REPAIR SYSTEM for pipe leak repair that has won wide acceptance.

The STOP IT PIPE REPAIR SYSTEM provides a method to restore product flow in minutes, instead of the many hours required to cut out a pipe section, fabricate and install a new piece. STOP IT’s properties allow it to stop leaks in about 30 minutes and it has a wide range of chemical resistance. STOP IT can be found in most industries including refineries, chemical plants, pulp and paper mills and power plants.

The introduction of Stop It HPT for leak repair and rehabilitation of high-pressure pipe systems operating up to 2,000 psi has expanded the company’s leak sealing capabilities far above the 400 psi of the original Stop It Pipe Repair System.

The company’s recent alignment with Amcorr’s VISCOTAQ line of anti-corrosion sealants allows us to provide the latest technology for use in the corrosive environments our customers operate in.

InduMar is continuously developing new solutions for leak and cor-rosion problems. Let us know about your leak or corrosion problems.

CONTACT INFORMATION3355 W Alabama St Suite 110 Houston, TX 77098 [email protected]

Stopit HPTM repair of methane pipeline operating at 550 psi


SPONSORED CONTENT HYDROCARBON PROCESSING | TERMINALS AND STORAGE 2013 T–97

CB&I PROVIDES GREATEST EXPERIENCE OF ANY TANK CONSTRUCTION COMPANY

CB&I is the most complete energy infrastructure focused company in the world and a major provider of government services. Drawing upon more than a century of experience and the expertise of approximately 50,000 employees, CB&I provides reliable solutions while maintaining a relentless focus on safety and an uncompromising standard of quality.

CB&I combines proven process technology with global capabilities in engineering, procurement and construction to deliver comprehensive solutions to customers in the energy and natural resource industries. With premier process technology, proven EPC expertise, and unrivaled storage tank experience, CB&I executes projects from concept to completion.

With over 46,000 tanks built in more than 100 countries, CB&I has accumulated more storage design and construction experience than any other organization in the world. In addition to being a leader in engineer-ing, procurement, fabrication and construction of storage tanks, we have also designed and built more than 100 storage terminals. We have the capability to design and install pipelines for these facilities, as well as other ancillary equipment.

Many customers draw upon this knowledge and extensive construc-tion experience early in a project’s development, enabling us to provide project-specific solutions that deliver maximum long-term value, lower up-front costs, and shorter schedules.

Safety is a core value at CB&I, and we are proud to have one of the best safety records in the industry. Throughout our organization, every employee worldwide is committed to safe work practices. Our award winning safety program promotes a culture of involvement and dedica-tion with a goal of zero incidents for everyone involved in our projects.

CONTACT INFORMATION2103 Research Forest DriveThe Woodlands, TX 77380 USATel: +1 832 513 1000Fax: +1 832 513 [email protected]

CB&I

www.ConstructionBoxscore.com

COMING SOON

The Redesigned and Enhanced ConstructionBoxcore Database• Details for thousands of active, global construction projects in the refi ning, petrochemical, gas processing and LNG industries

• Streamlined Quick Search Functionality that fi lters projects by region, project type, time and company

• Contact information for key personnel

• The Boxscore Update e-newsletter sent weekly with new projects and updates

• NEW: Customized Reports

Logon to www.ConstructionBoxscore.com and discover how the HPI’s most trusted source of construction data just BECAME EVEN MORE POWERFUL!

For more information, contact Lee Nichols, Data Director, at +1 (713) 525-4626

Welcome to the NEW Construction Boxscore Database


Refining DevelopmentsT. TEMUR, M. HAKTANIR, F. UZMAN, and

M. KARAKAYA, Turkish Petroleum Refineries Corp.

(Tüpraş), Kocaeli, Turkey; and A. K. AVCI, Boğaziçi University, Istanbul, Turkey

Optimize vacuum ejector operations

Vacuum distillation unit (VDU) designs generally involve jet ejectors due to their low capital and operating expenses, high reliability, durability and simplicity. Once the ejector is installed and commissioned, it may never require any maintenance as long as it is operated within its design limits. The major draw-backs associated with using ejectors are the dramatic perfor-mance losses with deviations from operating condition limits and low thermal efficiency.1

Ejector models, found in the literature, were developed for design purposes that estimate the cross-sectional areas of the converging and diverging sections or their respective ratios, and for determining intermediate pressures and velocities.1–4

Upon validation against field measurements, these models can be used to evaluate the performance of installed ejectors and enhance operating efficiency.

Case history. Simulation and evaluation of ejectors and field trials have focused on a wet-fuel vacuum unit with precondens-ers at Tüpraş’ Izmit refinery. The ejector system comprises of three-stage parallel ejectors, a precondenser, an intercondenser and an aftercondenser, as shown in FIG. 1. The ejectors at each stage are designed to accommodate one-third and two-thirds of the load. Motive steam to the ejectors is supplied directly from the low-pressure steam header at a nominal pressure of 3.5 kgf/cm2. Due to numerous branches from the refinerywide steam header, instantaneous pressure variations are unavoidable; thus, the ejectors are subjected to unregulated motive steam flow with pressure generally above the ejector design value of 3.5 kgf/cm2. A practical procedure for estimating cost savings via regulation of the steam pressure is evaluated in the presented case study. Annual savings for a 44,500-bpd vacuum unit is esti-mated to reach $526,000.

Procedure. Prior to the field trials, theoretical evaluations of the ejector performance were conducted to screen the conditions at which the ejector will function properly. The principal condition is that motive steam reaches supersonic velocities (Ma > 1) in the nozzle throat.5 As shown in FIG. 2, the evaluation task is ini-tiated by validating a mathematical model that predicts—along the length of the ejector—the pressure and velocity profiles, either of the motive steam or of its mixture with the entrained vapor from the vacuum tower. The validation step is based on comparing steam consumption by the actual ejector system and

those calculated by the model, which is easy to apply and capable of accurate predictions. If the discrepancies between the model results and time-averaged process data are high, then the model must be revised by progressive relaxation of certain assumptions.

Basic model. The one-dimensional (1D) model assumes that mixing the motive steam and entrained vapor occurs at con-stant pressure.1 The other major assumptions forming the basis for the calculations are:1

• The mixing pressure is equal to the saturation pressure of steam at Tmotive.

• Flow along the ejector is adiabatic.• Expansion of motive steam in the nozzle is isentropic.• Isentropic expansion coefficient is constant.• Motive steam and the mixture follow ideal gas behavior.• Velocity of the entrained vapor is zero.• Outlet velocity of the mixture is negligible.The steam flowrate through each ejector is calculated as:

(1)

mmotive kg s A1 Pmotive

RTmotive

n

12

1 –1

PI PI

PIPI

PIPI

Offgas

BA

Surge drumH6

Barometricseal drum

Vacuumdistillation

column

Steam header

Proposed pressurecontrol system

1st-, 2nd- and 3rd-stage ejectors

Precondensers C D E F

FIG. 1. Process flow diagram of the three-stage ejector system.


Refining Developments

where Tmotive is given by the Antoine equation (Eq. 2) for tem-perature:

(2) Tmotive C⎡⎣ ⎤⎦= 42.678− 3,892.7

ln Pmotive 1,000( )−9.486

⎡

⎣

⎢⎢⎢

⎤

⎦

⎥⎥⎥−273.15

The compression ratio, γ, and the nozzle efficiency, ηn, are assumed to equal to 1.31 and 0.85, respectively.1,6 The cross-sectional areas of the nozzle throats differ from stage to stage, but are identical for a pair of parallel ejectors. Position indica-tor 1 (FIG. 3) denotes the midpoint of the nozzle throat, while 2 denotes the nozzle outlet. When comparing the model results with the actual consumption data for validation, consistency leads to the next step. However, high discrepancy requires modifying the existing model, which is beyond the scope of this article.

Critical throat velocity. The ratio of the cross-sectional area of the nozzle outlet (position 2 in FIG. 3) to that of the nozzle throat (position 1) can be expressed as:

(3)

A2

A1

=1

Ma22

2γ +1

1+γ −1

2Ma2

2⎛⎝⎜⎜⎜

⎞⎠⎟⎟⎟

⎡

⎣⎢⎢

⎤

⎦⎥⎥

γ+1( )/ γ –1( )

Eq. 3 can be solved for the Mach number at the outlet (Ma2) since both the throat and outlet diameters are known from the equipment data sheets. The velocities of the motive steam into the ejectors can be calculated using the flowrates obtained from Eq. 1. The throat velocities are then interpolated linearly between the values at the ejector inlet and nozzle outlet. If Ma1, the throat velocity, is less than the critical value of 1, the ejector shows a poor performance, which leads to loss of vacuum at the top of the tower.

Minimum steam pressure. Once the relationship between the motive steam pressure and nozzle throat velocity is estab-lished, in practice, it remains a trial-and-error procedure to determine the minimum motive steam pressure at which the critical throat velocity can be reached, and to realize the opera-tional cost savings if steam from the header had been regulated prior to the ejectors. Note: A safety margin of Ma = 0.05 is al-lowed to prevent potential malfunctioning of the ejectors dur-ing the field trials.

Optional rigorous path. The discussed model is based on 1D gas dynamics. The obtained results are expected to shed light on how the actual ejector system configuration should be tested in the field. Conversely, when verified by experimental data, a computational fluid dynamics (CFD) model of a system is a very useful tool for testing and analyzing complex geomet-ric designs under a variety of operating conditions.

Apart from the basic model, the CFD model of the ejector is constructed to predict the nozzle throat velocity by provid-ing a detailed CAD drawing, and the inlet steam velocity and outlet pressure as boundary conditions. In the ejector, the velocity and pressure profiles of motive steam and entrained vapors can be obtained over the entire domain by solving the compressible turbulent Navier-Stokes equations. Depending on the complexity of the geometry, the grid resolution and the underlying physics (e.g., high Mach number flows), solution methods for these sets of equations may demand too much

Calculate steamconsumption by theejectors for various

Pmotive (basic model)

Read the steam headerpressure, Pmotive

Pmotive = Pheader

Do the modelpredictions compare

with averageprocess data?

Calculate total steamconsumption at Pmotive

by the ejectors

Calculate the inlet velocityof motive steam into the

3rd-stage ejectors for Pmotive

Determine the nozzle throatvelocity in the 3rd-stage

ejectors for Pmotive (CFD model)

Stop!Minimum Pmotive has been

reached. Further reduction inmotive steam pressure can

lead to loss of vacuumat the tower top

Is throatvelocity

greater thanMa = 1.05?

Determine the nozzle throatvelocity in the 3rd-stage

ejectors for Pmotive (Basic model)

No

Yes

Yes

Field trial at this Pmotive isallowed. Calculate actual

steam consumption by theejectors

Calculate the operating costsavings by reducing the motive

steam pressure

YesNo

Optional

No

Revise the ejector model,e.g., raise the hierarchy

Modelvalidation

Throatvelocity

estimation

Reduce the currenttrial pressure:

Pmotive = Pmotive–0.1 kg/cm2

Is throatvelocity

greater thanMa = 1?

FIG. 2. Procedure for determining the cost savings by working at the minimum possible motive steam pressure of the ejectors.

201

Veloc

ity, M

aPr

essu

re, kg

f/cm2

2

3 4 5

2100.9

1.01.11.21.31.4

Entrainedvapor

Motive steam

1234

3Location along the ejector

4 65

FIG. 3. The basic model results for the pressure and velocity profiles along the ejector H6 for different motive steam pressures.



computational power and time.7 However, pertaining to the ex-perience of the authors, any ejector system can be dealt within no more than a few seconds, using a now-standard desktop PC with 2 GB of RAM.

Case study. The time-averaged-process data can be used to validate the model. Steam header pressure and total consump-tion trends from Jan. 1, 2012, to Aug. 30, 2012, are shown in FIG. 4. For an average pressure of 4 kgf/cm2, the model forecasts the actual consumption rate within 4% error. On the day of the field trial, the header pressure over a 3-hour period was 3.7 kgf/cm2, which equates to a total steam consumption rate of 13.5 tph. The model prediction was 7.5% off the actual consump-tion rate, which is acceptable, considering the numerous as-sumptions made during development and possible orifice-flow measurement errors. TABLE 1 summarizes the model findings and actual figures.

The critical velocity at the nozzle throat of ejector H6 (FIG. 1) was estimated using Eq. 3 and the iterative procedure listed in FIG. 2 for motive steam pressures of 3, 3.5 and 4 kgf/cm2. The Mach number, which was fairly unaffected by reductions in pressure (see FIG. 3), was above the critical value of 1. Thus, it is possible to operate the ejectors at pressures around their minimum design value of 3 kgf/cm2. Decreasing the pressure below 3 kgf/cm2, however, led to throat Mach numbers less than the critical velocity, and, thus, a loss of vacuum at the top of the tower.

Findings of the basic model were verified by the CFD mod-el, and the results are shown in FIG. 5 for motive steam pressures of 4, 3.5, 3 and 2.8 kgf/cm2. Nozzle throat velocities for the first three pressures that were above the critical Mach number of 1—while going below 3 kgf/cm2—pulled the velocity below sonic levels. The CFD model considered only motive steam flow through the ejector, since reaching supersonic speeds at the nozzle throat is the most important criterion for a success-ful ejector operation.5 The suction section was not included in the computational domain.

Field trials. In the Tüpraş refinery, the system considered items A through F, as shown in FIG. 1. The gate valves admit steam to the ejectors from the header, and the valves cannot be used to regulate steam flow during normal operation. To prevent an irreversible upset of the vacuum system, only the valve (item F in FIG. 1) that is associated with the smaller of the third-stage ejectors (H6) was restricted with extreme care to reduce the steam pressure in accordance with the results of the basic and CFD models. All the other five ejectors were subjected to steam at the header pressure (3.7 kgf/cm2 on average). Upon the reduction from 3.7 kgf/cm2 to 3 kgf/cm2, the lower limit of safe operation, the column vacuum was actually increased by 3 mm Hg (FIG. 6), while total steam consumption rate de-creased by 0.2 tph. The vacuum increase is explained by the relief of the aftercon-denser loading as the amount of steam condensed is reduced.

Although the motive steam pressures below 3 kgf/cm2 were anticipated by the models to be detrimental to operation, the system was technically analyzed by the engineers and opera-tors. It was concluded that drawbacks resulting from vacuum loss for a very short time would be minimal. Therefore, the steam pressure through ejector H6 was further reduced to 2.4 kgf/cm2 by tighter restriction of valve F, which, however, led to a dramatic 16% increase in the column top pressure from 45 mm Hg to 52 mm Hg. Restriction on valve F was immedi-ately lifted, and the pressure increased back to 3 kgf/cm2. Once again, the model prediction for the system was correct.

Economic gain through regulation. From the model-ing studies and field trials, it is apparent that regulating the pressure of motive steam leads to considerable reduction in

3.5 12

13

14

15

16

11/3/2011 12/23/2011 2/11/2012 4/1/2012 5/21/2012 7/10/2012 8/29/2012 10/18/2012

Steam

head

er pr

essu

re, kg

f/cm2

Tota

l ste

am co

nsum

ption

, tph

3.63.73.83.94.04.14.24.34.44.54.64.74.84.95.0

FIG. 4. 2012 trends of steam header pressure and total steam consumption by the ejector system.

FIG. 5. Velocity profiles in the ejector and nozzle throat (insets) obtained with the CFD model for various motive steam pressures.

TABLE 1. Comparison of model results and actual steam consumption rates

Motive steam

pressure, kgf/cm2

Model results for consumption at ejector stage, tph

Model results for total

consumption, tph

Actual consumption,

tphError,

%

1 2 3

3.7 2 4.2 7.3 13.5 14.6 7.5

4* 2.1 4.5 7.9 14.4 14.9 4* Year-long (2012) average of steam header pressure



consumption without impacting the vacuum system. From FIG. 1, installing a control valve on the steam header before it branches off to the ejectors is necessary to maintain pressure at 3 kgf/cm2.

The economic consequence of such a commonplace opera-tion is immense compared to its simplicity. Operational cost savings for the vacuum system by the reduction and regulation of the motive steam pressure is summarized in TABLE 2. Con-sumption rates at 3.5 kgf/cm2 are values at the minimum design pressure, and are obtained from the equipment data sheets; whereas, rates at 3 kgf/cm2 are calculated using the basic mod-el. The projected reduction in steam consumption for the ejec-tor system under consideration is 1.67 tph, and its economic payoff is $526,000/yr. Remember: This systematic demon-stration of the concept that even the most steadfast of the refin-ery equipment should be evaluated for potential performance increases and cost reductions. When the evaluation task in the field is expensive and poses risks to unit operations, even the simplest process models can assist engineers or unit operators develop a safe procedure for the specific evaluation.

NOMENCLATUREA1 Cross-sectional area at the nozzle throat, m2

A2 Cross-sectional area at the nozzle outlet, m2

mmotive Motive steam flowrate through the ejectors, kg/sMa1 Mach number at the nozzle throatMa2 Mach number at the nozzle outletPmotive Pressure of motive steam, kgf /cm2

R Universal gas constant, kJ/kg-KTmotive Temperature of motive steam, °C� Compression ratio�n Nozzle efficiency

LITERATURE CITED 1 El-Dessouky, H., et al., “Evaluation of steam jet ejectors,” Chemical Engineering and

Processing, June 2002, Vol. 41, pp. 551–561. 2 Keenan, J. H. and E. P. Neumann, “A simple air ejector,” Journal of Applied

Mechanics, Vol. 9, No. 2, pp. A75–A81. 3 Huang, B. J., et al., “A 1D analysis of ejector performance,” International Journal of

Refrigeration, May 1999, Vol. 22, pp. 354–364. 4 Aly, N. H., A. Karmeldin and M. M. Shamloul, “Modeling and simulation of steam

jet ejectors,” Desalination, January 1999, Vol. 123, pp. 1–8. 5 Lieberman, N. P. and E. T. Lieberman, A Working Guide to Process Equipment,

2003, 2nd Ed., McGraw-Hill, New Jersey. 6 Eames, I. W., S. Aphornaratana and H. Haider, “A theoretical and experimental

study of a small-scale steam jet refrigerator,” International Journal of Refrigeration, June 1995, Vol. 18, pp. 378–385.

7 Nazarov, M. and J. Hoffman, “Residual-based artificial viscosity for simulation of turbulent compressible flow using adaptive finite element methods,” Numerical Methods in Fluids, No: 10.1002/fld.3663.

TOLGA TEMUR is a senior process engineer and has been with Tüpraş since 2008. He is responsible for the crude units and vacuum processes. Mr. Temur also has experience in cracking and hydrodesulfurization processes. He holds a BS degree in chemical engineering and an MS degree in fuel and energy technologies from Boğaziçi University.

MERT HAKTANIR is a senior process engineer at Tüpraş and has four years of engineering and troubleshooting experience in catalytic cracking and crude/vacuum distillation units. His current interests include upgrading and conversion of heavy crudes and shale oils, and detailed kinetic modeling of complex reaction systems. Mr. Haktanir holds a BS degree in chemical engineering from Boğaziçi University, and MS degree

in chemistry from Purdue University. At present, he is pursuing a PhD in process and reactor design from Istanbul University.

FIRAT UZMAN joined Tüpraş in 2010, and is the senior R&D engineer with expertise and focus on modeling, control and optimization of refinery processes. His previous career experience includes tenure in the automotive industry. Mr. Uzman holds a BS degree in mechatronics engineering from Sabanci University, and is pursuing an MS degree in industrial engineering from Koç University.

MUSTAFA KARAKAYA, PhD, is an R&D engineer with Tüpraş. He received BS and MS degrees and a PhD in chemical engineering from Boğaziçi University. His professional and research interests include CFD design and analysis of catalytic reactors, development of GTL catalysts, and modeling and optimization of hydrotreating processes.

AHMET K. AVCI, PhD, has received BS and MS degrees and a PhD in chemical engineering from Boğaziçi University. From 2003 to 2005, he worked as the R&D manager for Procter & Gamble. In 2006, he joined Boğaziçi University as a full-time faculty member. Assoc. Prof. Avci’s research interests are focused on catalytic hydrogen and synthesis gas production and conversion technologies, and on intensification of catalytic

reactors by microchannel technology. He is the leader of numerous projects funded by governmental research institutes and the industry and is the author of more than 20 articles published in international journals.

Pmotive = 2.4 kgf/cm2

Pmotive = 3 kgf/cm2

Pmotive = 3.7 kgf/cm2

2:52 pm

Vacu

um co

lumn t

op pr

essu

re, m

m H

gTo

tal s

team

cons

umpt

ion, t

ph

4042

44

46

48

50

52

54

13.0

14.0

15.0

13.5

14.5

15.5

3:21 pm 3:50 pm 4:19 pm 4:48 pm 5:16 pm 5:45 pm 6:14 pm 6:43 pm 7:12 pm 7:40 pm

FIG. 6. Variation of total steam consumption by the ejectors and the vacuum column top pressure by the pressure of motive steam into ejector H6.

TABLE 2. Operational cost savings upon regulation of the motive steam pressure

Motive steam

pressure, kgf/cm2

Steam consumption at ejector stage, kg/h

Total reduction,

tph

Cost savings,

$million/yr

1 2 3

3.51 1,896 3,986 7,026

32 1,651 3,471 6,117

Reduction 245 515 909 1.67 0.5263

1 Design value2 From the basic model3 Based on the year-long average of cost of steam generation, $36/t

Expert Engineering. Proven Results.

We are never satisfied.

At Waukesha Bearings Corporation™, our forward thinking team is committed to leading the markets we serve through

application knowledge, superior technology and operational excellence.

Our bearing and seal solutions are engineered to meet the unique needs of our customers, giving us the broadest application experience in oil & gas, power generation, and industrial markets. Our global businesses are customer focused –

in everything we do – by delivering excellence in technology, product performance, and customer service and support.

Count on our team. We’re there when you need us.

www.waukeshabearings.com/HPSelect 74 at www.HydrocarbonProcessing.com/RS

33RD ANNUALInternational Operating Conference & Trade Show

JUNE 3-5, 2013Hilton Americas-HoustonGeorge R. Brown Convention CenterHOUSTON, TEXAS

KEYNOTE SPEAKER

Cal Ripken, Jr.• Baseball’s Iron Man and Hall of Famer

Retired from baseball in October 2001 after

21 seasons with the Baltimore Orioles, Cal

Ripken, Jr. was entered in the record books

as one of only seven players in history to

achieve 400 home runs and 3,000 hits.

Cherished by fans around the globe as

baseball’s “Iron Man,” in 1995 he broke Lou Gehrig’s record for

consecutive games played, and voluntarily ended his streak in 1998

after playing 2,632 consecutive games. This is regarded as one of

the single greatest accomplishments in sports history. Ripken’s name

has become synonymous with strength, character, endurance, and

integrity. His philosophy of working hard, playing with passion,

and enjoying the game has made a tremendous impact on the sport

and on fans everywhere. ★

[email protected] | +1-703-875-2011 | www.ilta.org

CONFERENCE SESSIONS INCLUDE:

Tank Overfill Protection: Evaluating the Revised API 2350 Standard

Mechanical Integrity Programs: The Case for a Proactive Approach

The Panama Canal Expansion and Other Critical Issues for U.S. Ports

America’s Second Industrial Revolution: The Renaissance of U.S. Chemical Manufacturing

Mitigating Cyber Threats: Tools for Improving Computer and Network Security

Ensuring Due Diligence in Dealing with Vacuum Truck Hazards

Develop your technicalexpertise in terminal operations by attending focused presentations and workshops.

Plus!13 conference presentations will be conducted in Spanish



Safety/Loss Prevention

J. THARAKAN, Suncor Energy Products,

Calgary, Alberta, Canada

Flare header failure: An investigation

A 60-in. flare header suffered three identical failures with buckling and cracking at the circumferential weld at the bottom of the pipe. The failures are believed to have originated during hot reliefs.

The flare header handles liquid, vapor and gas intermittently relieved from various process units. The flare header runs hori-zontally for a length of 3 km and has a built slope of 1 in 500 for free draining. The hottest relief is from the coker unit, which could be at a maximum temperature of 815°F.

The flare header was designed to ASME B 31.3, non-severe cyclic service with a design pressure of 50 psig at 815°F. It was fabricated using 0.375-in. thick carbon steel plate with non-post-weld-heat-treated welds. The header was also heat traced and insulated.

FIG. 1 depicts the failure locations. All failures were with-in about 60 ft on either side of the tie-in from the sulfur unit (sulfur lateral).

FIG. 2 shows a typical buckling and cracking failure at the bot-tom of the circumferential weld on the flare header.

Points to ponder. At the time of investigation, the failed areas were enclosed with leak containment devices (LCDs) installed to arrest leaks on a temporary basis. Therefore, closer observa-tion of the failure wasn’t possible.

Buckling occurs due to compressive stress (or compressive strain) and cracking is caused by tensile stress. Therefore, it ap-peared strange how both could occur at the same location. An-other boggling question was which occurred first, the buckling or the cracking?

The welds are, in general, stronger than the pipe. Why would the girth weld area buckle, as opposed to the pipe itself?

Why were all the failures occurring at the vicinity of the sul-fur lateral? The buckling at the bottom section of the pipe could

be associated with a bending of the pipe in the vertical plane. The pipe is designed to move in the horizontal plane during thermal growth. Why should it move in the vertical plane?

Movement of the pipe in the vertical direction can result if there is a temperature difference between the top and bot-tom of the pipe. Therefore, it was suspected that there could be some insulating deposit at the bottom of the flare header near the sulfur lateral.

Onstream inspections. Gamma ray scanning ruled out the presence of internal deposits in the flare header. Scanning re-sults from the welds and the pipe near the sulfur lateral elimi-nated environmental cracking and thinning.

Visual inspection of the 60-in. header revealed:• The pipe developed an ovality due to a support reaction

at the vicinity of the saddle supports

Verticalrestraint Guide

From cokerunit

60-in. flareheader

110 ft

Axialrestraint

Slidingsupport (typ.)

Failu

re-2

To flareknockout pot

Sulfurrelief

Failu

re-3

Failu

re-1

FIG. 1. Failure locations: All failures were within about 60 ft on either side of the tie-in from the sulfur unit (sulfur lateral).

FIG. 2. Typical buckling and cracking failure at the bottom of the circumferential weld on the flare header.

FIG. 3. Evidence of vertical movement of the pipe.



• The pipe was not touching one of the bottom supports (1-in. gap), due to a permanent deformation in the vertical plane

• The lone Y- stop that prevented upward movement of the line was forced open (see FIG. 3)

• There was no ovalization at the long radius bends• There was no sign of excessive axial or lateral movement• There were no fretting marks on supports, that could be

associated with vibration.The first three observations led to the conclusion that the

flare header must have undergone thermal movement in the vertical plane.

Design observations. Several design features were observed during the investigation. The saddle support for the 60-in. flare header had an angle of contact of 72°. This is less than the mini-mum angle of contact of 120° for saddles for horizontal pressure vessels. The flare header was not designed for hydro-fill condi-tions. Some pipe spans that exceeded 40 ft, coupled with defi-ciency in saddle design, caused high local stress and ovalization of the pipe near supports, as confirmed with stress analysis.

The pipe thickness (t) was too thin for its outside diameter (D), i.e., it had a large D/t ratio of 160, as against the industry norm of ≤120. In FIG. 4, it can be seen that when the D/t ratio increases, the allowable strain limit decreases.

Strain limit. When a pipe is bent by the application of an ex-ternal moment, it tends to develop changes in cross-section. The outer radius develops flattening and inner radius kinks inward (buckling).

Strain limit is often used to assess the bending capability of pipe. As shown in FIG. 4, the strain limit decreases with an increase in the D/t ratio. The strain referred to here is the mechanical strain that produces stresses. Free thermal expansion produces thermal strain without stressing the material. In a piping system, restraints always limit free thermal expansion; therefore, some mechanical strain is also induced during thermal expansion.

Gresnigt’s equations are the basis of FIG. 4.1

Єc = 0.5 � t/(D-t) – 0.0025 for (D-t)/t < 120Єc= 0.2 � t/(D-t) for a (D-t)/t ≥ 120

With fast response TDLS technology that monitors closer to the burn, the SERVOTOUGH LaserSP’s unique fl ue gas analysis solution reduces fuels costs and CO2 emissions. You can’t get a faster response.

The SERVOTOUGH SpectraScan is a breakthrough in measurement speed for light hydrocarbons C1-C5+. A stand-alone solution with superior technology and benefi ts, the SpectraScan changes everything.

AIMHIGHERHIT YOUR FURNACE COMBUSTION TARGETS WITH THE SERVOTOUGH LASERSP

EFFECTIVE COMBUSTION CONTROL IS CRITICAL TO EFFICIENT ETHYLENE PRODUCTION.

A MEASURABLE ADVANTAGES E R V O M E X . C O M

FIND OUT HOW YOU COULD IMPROVE THE EFFICIENCY OF ETHYLENE PRODUCTIONDOWNLOAD THE WHITE PAPER FROM:

SERVOMEX.COM/LASERSP OR TO REQUEST ONE, EMAIL: [email protected]

LEARN MORE ABOUT THE REVOLUTION IN HYDROCARBON GAS ANALYSISDOWNLOAD THE WHITE PAPER FROM:

SERVOMEX.COM/SPECTRASCANOR TO REQUEST ONE, EMAIL: [email protected]

The NEW SERVOTOUGH SpectraScan

MOVE OVER GCTHE FUTURE OF ON LINE GAS ANALYSIS IS HERE

-0.01000

0.00000

0.01000

0.02000

0.03000

0.04000

0.05000

0.06000

0.07000

0 50 100 150 200

Єc

D/t ratio

FIG. 4. D/t ratio vs. strain limit, Єc.



�107

D/t is not the lone parameter limiting the bending capacity of pipe. The other factors are:

• Nonhomogeneity or the presence of imperfections in the material

• Initial out-of-roundness• Loading conditions• Residual stress• Strength of the material in the longitudinal and circum-

ferential directions• Shape of the stress–strain diagram.

Thermal bowing. The flare header is designed for flexibil-ity at 815°F. As per the design, it is expected to move only in the horizontal plane while undergoing thermal expansion. A movement in the vertical plane has to counter the gravity loads; however, the several conditions have led to thermal bowing of the flare header in the vertical plane:

• Occasional temperature difference between the top and bottom half of the horizontal flare header

• Ovalization of the pipe, which reduced the stiffness of the pipe in the vertical direction, compared to the horizontal direction

• To accommodate growth, the pipe tends to bend in the di-rection of least stiffness.

The reason for the circumferential thermal gradient is not well understood. Within the horizontal flare header, there could be a two-phase flow of hot fluid, with the film-heat transfer co-efficient being different for the liquid flow and the vapor flow. This difference can cause a temperature gradient between the bottom and top of the pipe. Another possibility is the intermit-tent and partial flow of liquid, hotter or colder than the mean line temperature; it could create a circumferential thermal gradient.

FIG. 5 shows thermal bowing of a pipe due to circumferential thermal gradient. The pipe will bend with outer curvature at the hotter region. The nature and extent of stresses generated is dependent on the boundary condition. In this example, the vertical restraint opposes free thermal bowing. The region of the pipe where actual displacement is less than needed for free thermal expansion would be in compression and if reversed, in tension. Therefore, the hotter half of the pipe will be under compression and the colder half in tension.

Initial conclusions. The preliminary investigation conclud-ed that the failure was initiated by local buckling at zones of compressive strain when thermal bowing occurred in the flare header. The vertical restraint amplified the stresses at the re-gion of Failure 1. At failure locations 2 and 3, the sulfur lateral restrained the rotation and lifting of the 60-in. header, thereby increasing the stresses.

The large pipe D/t ratio is the root cause of local buckling. From FIG. 4, the strain limit = 0.00126 for the flare header with a D/t ratio of 160. This translates to a stress of 35.3 ksi (1 ksi = 1,000 psi), which is lower than a yield stress of 38 ksi. Imper-fections and residual stress at the circumferential weld lower the strain limit, thus explaining how all three failures occurred at the circumferential welds.

Hoop tensile stress opposes inward buckling of the pipe. The flare header operating pressure does not exceed 10 psig and this low value added negligible hoop tensile stress to counter buckling.

The cracking occurred after the buckling damage. For crack-ing, tensile stress is required, and this must have resulted dur-ing straightening of the pipe when thermal gradients receded or when reversed.

About 150 ft on either side of the sulfur lateral was deemed as the only area where all the following conditions required for local buckling co-existed:

• High temperature due to downstream coker unit relief• Restraints that opposed thermal bowing• Large D/t ratio• Pre-existing ovality due to large local stresses near saddle

supports.

Repairs carried out. After thorough evaluation, repair strat-egy improvement began. Key improvements included:

Cold side–tensile stress

Verti

cal

restr

aint

Hot side–compressive stress

Sliding support

FIG. 5. Thermal bowing of pipe due to circumferential thermal gradient.




• A 300-ft section encompassing the failure was replaced with a thicker pipe

• The D/t ratio for the replacement section was 96• The saddle supports for the replaced section were rede-

signed with an angle of contact of 120°• The vertical restraint was modified to a sliding support• Skin thermocouples were installed at the top and bottom

of the pipe• The new welds were post-weld-heat-treated to reduce re-

sidual stress and to safeguard against environmental cracking.

Metallurgical examination. A close inspection of the flare header section removed for metallurgical inspection is shown in FIG. 6. Metallurgical observations included:

• Metallurgy of the pipe and weld were verified and found to match with the original design specification

• No weld defects were detected; two of the failures were on shop welds and one was on a field weld

• No sign of fatigue or environmental cracking was found on the specimens examined

• The bottom portion of the pipe, at the weld, buckled in-ward and the cracks originated at the buckled area

• The cracking was due to ductile overload• The crack originated at the toe of the weld from the OD

surface of the pipe at the buckled region.Skin temperature readings taken from the top and bottom

of the pipe after replacement revealed a circumferential ther-

www.paqell.com

Award winning technology for sulphur removal and sulphur recovery from gas streams. Take the quick scan at www.paqell.com/scan and receive instant feedback about how THIOPAQ O&G technology matches with your desulphurisation requirements!

GAS DESULPHURISATIONWITH THIOPAQ O&G: STABLE & COST EFFECTIVE

015-0047 Adv Hydrocabon 180x125mmn 13112012 v4 def.indd 1 13-11-12 16:48

FIG. 6. Failed section of 60-in. flare header.




mal gradient of 150°F. In general, the pipe is hotter at the bot-tom; however, in some instances, the temperature gradient is reversed, with the top of the pipe being hotter.

Discussions. Vertical movement of the pipe causing lift-off from support will add gravity loads to the pipe. Thermal bow-ing will negate the free draining capability of the line, leading to localized liquid pooling and associated issues. Thermal bowing is not common with process plant piping, and when it occurs, it is a difficult problem to correct. Most failures due to thermal bowing are fatigue cracking at the circumferential welds. Buck-ling due to thermal bowing is extremely rare due to installing pipes with favorable D/t ratio. The 60-in. flare header had both a very large D/t ratio and a loading/displacement condition that increased vulnerability to buckling.

When large lateral displacements are imposed on piping, failure generally manifests as localized buckling. Buckling is a failure due to instability and it causes process of achieving equilibrium between external loads, internal resistance and boundary restraint. Strain-based design is typically adopted for displacement controlled designs. Examples are subsea piping or buried lines with large ground movements.

In a piping flexibility analysis, the displacement stress range is compared with the allowable stress range. This is essentially a check against potential fatigue failure due to cyclic tensile stress. Compressive stress or strain limit checks are not part of a piping flexibility analysis. Piping stress analysis softwares

treat pipe as a beam and cannot predict local buckling of the shell elements.

Large D/t ratio also increases susceptibility to failures due to acoustic induced vibration (AIV). AIV is caused by high sound pressure levels inside flare headers during significant relief scenarios. AIV failures typically develop at small bore tie-ins to the flare header.

Findings. If the D/t ratio of the pipe exceeds 100, these pre-cautions apply:

• When the pipeline is subjected to large bending mo-ments, external pressure or axial compression, strain-based de-sign/buckling assessment using finite element analysis (FEA)should be performed

• Equations for stress intensity factors given in ASME B31.3 are valid only for D/t ≤ 100

• During flexibility analysis, corrected stress intensity fac-tors estimated through FEA should be used

• Flare headers should be designed with a dead load that includes one quarter full of liquid

• Saddle supports for piping with a large D/t ratio require design considerations like pressure vessel saddles (Zick analysis)

• The D/t ratio for flare headers should be less than 120.

LITERATURE CITED 1 Gresnigt, A. M. and R. J. Van Foeken, “Local buckling of UOE and seamless steel

pipes,” 11th International Offshore and Polar Engineering Conference, Stavanger, Norway, June 2011.

JULY 30–31, 2013Norris Conference Centers – CityCentre Houston, Texas

GulfPub.com/GTL

Hydrocarbon Processing Introduces the Inaugural Gas-to-Liquids Technology Forum Gulf Publishing Company’s inaugural Gas-to-Liquids (GTL) Technology Forum will investigate the technology and trends at work as GTL usage and projects become increasingly popular. The two-day technology conference will feature industry keynotes and presentations from GTL technology experts.

GTL is poised to become an increasingly important part of the North American energy industry. As the natural gas boom in North America continues and new technologies emerge to reduce costs, company interest is increasing—and so is investment. Project announcements and planning from industry innovators like Sasol, Shell and BP are beginning to ramp up in this energy niche.

GTL Technology Forum is your chance to connect with top operators and technology leaders from across the global hydrocarbon processing industry. Sponsorship at this event will benefi t the following types of companies active in the GTL sector: GTL technology providers, GTL engineering companies, major and independent operators, catalyst companies, specialized equipment manufacturers, turbine specialists and others.

For more information, contact your local Hydrocarbon Processing sales representative or Bret Ronk, publisher, at +1 (713) 520-4421 or [email protected].

HP0513_GTL Hlfh.indd 1 4/15/13 10:01 AM

MARKETPLACE / [email protected] / +1 (972) 816-3534


Specialty Engineering–Static Equipment–Rotating EquipmentMetallurgical and

Materials LabField Service

Specialists in design, failure analysis, and troubleshooting of

static and rotating equipment

www.knighthawk.com

Houston, exasel:

Fax:



SURPLUS GAS PROCESSING/REFINING EQUIPMENT 25 MMCFD x 1100 PSIG PROPAK REFRIGERATION PLANT

28 TPD SELECTOX SULFUR RECOVERY UNIT1100 BPD LPG CONTACTOR x 7.5 GPM CAUSTIC REGEN

NGL/LPG PLANTS: 10–600 MMCFDAMINE PLANTS: 60–3300 GPMSULFUR PLANTS: 10–180 TPD

FRACTIONATION: 1000–25,000 BPDHELIUM RECOVERY: 75 & 80 MMCFD

NITROGEN REJECTION: 25–100 MMCFDMANY OTHER REFINING/GAS PROCESSING UNITSWe offer engineered surplus equipment solutions.

Bexar Energy Holdings, Inc.Phone 210-342-7106 Fax 210-223-0018

www.bexarenergy.com Email: [email protected]



1,000 GPM AMINE PLANT

1250 PSI AVAILABLE NOW

More info: www.gas-corp.com/amineplant

Gas Corpora on of America

800-762-6015 * [email protected] * www.gas-corp.com


Visit HydrocarbonProcessing.com

New Version! Now includes Plate Fin Exchangers. For the designer or operator in the petroleum and chemical industry. Using over 40 years of design and operating experience by the author, the program calculates units that enable the inexperienced technical user to determine accurate design or evaluate operations. For the design and evaluation of process equipment:

Price: $3,295

programs written in Basic and three

reservoir engineering programs are linked

menus. However, each program is self-contained and can be used independently.

for their usefulness in day-to-day operations that fall into six main categories:

comprehensive user manual.

Price: $250

SOFTWARE VIDEO BOOKS

Visit our website to see all that Gulf Publishing Company offers

www.GulfPub.comPhone: +1 713-520-4426Email: [email protected]

PROCESS TOOLSSimulations and Design Calculations for the Petro-chemical Processing Industry

PEPAC 1Reservoir Engineering Programs and Economic Evaluation Spreadsheets

PRaS



Turbomachinery Training Compressors Steam Turbines Gas Turbines

Performance Analysis, Evaluation, Troubleshooting Problem Resolution, Case Studies, Maintenance, Reliability

Train with the Best

EXPERT INSTRUCTORS WITH YEARS OF GLOBAL EXPERIENCE“I feel your course is very bene cial because of the technical content. It is one step more ad-vanced than a typical OEM course ...” Yuthana Preechalai, PTT Aromatics and Re ning, PLC, Rayong, Thailand

FREE PERFORMANCE ANALYSIS SOFTWARE• Maximize Plant Production & Reliability • Optimize Your Condition Based Equipment Reliability Program• Con rm OEM Performance Guarantee• Optimize Equipment Utilization

Denverwww. exwareinc.com

sales@ exwareinc.com1-724-527-3911

Flexware® Turbomachinery Engineers

®



CUSTOM

REPRINTS

For additional information, please contact

Foster Printing Service, the official reprint

provider for Hydrocarbon Processing.

Give yourself a competitive

advantage with reprints.

Call us today!

Call 866-879-9144

or [email protected]

Take advantage of

your editorial exposure.

REPRINTS ARE IDEAL FOR:

Product announcements

Sales aid for your field force

PR materials and media kits

Direct mail enclosures

Trade shows

Conferences



SPECIALIZE IN BASIC DESIGN/REVAMPING*

• Ammonia Plants • Methanol/H2-CO Plants• Re-rating Primary Reformers• Propane Dehydrogenation Plants• Acrylic Acid & Oxo-Alcohol Plants

KPI, Inc., Houston, Texas-USA; Mail: [email protected]: 281-773-1629; Fax: 832-565-9360; Web: www.kpieng.com

Providing Value through Process Expertise

* Including Process Design, Technology Evaluation,

Simulation Modeling, Troubleshooting, Revamp studies,

Project Cost Estimates & Customized Process Training

Kinetics Process Improvements


WABASH SELLS & RENTSBOILERS & DIESEL GENERATORSFAST EMERGENCY SERVICE

www.wabashpower.com800-704-2002

FAX: 847-541-1279 847-541-5600


Here is a three-part lecture series featuring Heinz Bloch, respected author and lecturer on process equipment reliability. Each program introduces important and proven recommendations that the operator can use to improve equipment reliability in their facility. Introduction and Overview, and

why there is no reliability without operator involvement

Common Misunderstandings with Equipment Reliability Impact

Avoiding Machinery Failures

Price: $595*

SOFTWARE VIDEO BOOKS

www.GulfPub.com

*Applicable tax, shipping and handling apply

The Operator’s Role in Achieving Equipment Reliability Series

Presented by Heinz Bloch

Detailed and up-to-date information for active construction projects in the refining, gas processing, and petrochemical industries across the globe | ConstructionBoxscore.com

HP113


Why Should You Filter Your Water?

The Best Engineered Water Filteration Solution Always Costs Less



ADVERTISERS INDEX / HydrocarbonProcessing.com

The first number after the company name is the page on which an advertisement appears. The second number, appearing in parentheses, after the company name, is the Reader Service Number. There are two ways readers can obtain product and service information:1. Go to www.HydrocarbonProcessing.com/RS. Follow the instructions on the screen, and your request will be forwarded for immediate action.

2. Go online to the advertiser's Website listed below.

Company Page RS# Website



ae Solutions ............................................................. 46–47 (165)www.info.hotims.com/45679-165

Aggreko ........................................................................ 41 (162)www.info.hotims.com/45679-162

Ametek Process Instruments ...........................................45 (164)www.info.hotims.com/45679-164

ARCA Regler GmbH ......................................................... 72 (169)www.info.hotims.com/45679-169

Avondale .......................................................................90 (77)www.info.hotims.com/45679-77

Axens ........................................................................... 116 (51)www.info.hotims.com/45679-51

Babbitt Steam Specialty Co. ............................................40 (160)www.info.hotims.com/45679-160

BETE Fog Nozzle .............................................................84 (58)www.info.hotims.com/45679-58

Borsig GmbH ................................................................. 19 (153)www.info.hotims.com/45679-153

Cameron ....................................................................... 52 (55)www.info.hotims.com/45679-55

Carboline Company ........................................................88 (171)www.info.hotims.com/45679-171

CB&I ...........................................................................T-96 (54)www.info.hotims.com/45679-54

Cudd Energy Services ......................................................71 (168)www.info.hotims.com/45679-168

Curtiss Wright Flow Control Company, DeltaValve ................... 68 (67)www.info.hotims.com/45679-67

Curtiss Wright Flow Control Company, Farris Engineering .........73 (56)www.info.hotims.com/45679-56

Dresser-Rand ................................................................. 67 (62)www.info.hotims.com/45679-62

Elliott Group ................................................................. 115 (52)www.info.hotims.com/45679-52

Emerson Process Management, Fisher ............................. 33 Emirates .........................................................................13 (68)

www.info.hotims.com/45679-68Exxon Lubricants & Specialities ........................................51 (71)

www.info.hotims.com/45679-71Flexitallic LP ....................................................................5 (93)

www.info.hotims.com/45679-93

GE Measurement & Control, Inspection Technology ........... 26 (155)www.info.hotims.com/45679-155

GE Oil & Gas, Surface Pumping Systems ...........................34 (86)www.info.hotims.com/45679-86

Gulf Publishing Company Construction Boxscore .............................................T-98 Events—GTL ............................................................. 109 Events—IRPC ...........................................................6–7 Marketplace ........................................................ 110–112 Hermetic Pumpen GmbH ................................................ 32 (158)

www.info.hotims.com/45679-158Hoerbiger ......................................................................30 (156)

www.info.hotims.com/45679-156HTRI .............................................................................. 22 (154)

www.info.hotims.com/45679-154Hydro, Inc. .....................................................................20 (83)

www.info.hotims.com/45679-83Hytorc ........................................................................... 37 (54)

www.info.hotims.com/45679-54Idrojet ........................................................................... 77 (170)

www.info.hotims.com/45679-170ILTA ............................................................................. 104 (59)

www.info.hotims.com/45679-59InduMar Products ........................................................ T-95 (172)

www.info.hotims.com/45679-172Inpro/Seal, A Waukesha Bearings Business .....................103 (74)

www.info.hotims.com/45679-74John Zink Company ....................................................... 27 (80)

www.info.hotims.com/45679-80Johnson Screens ............................................................ 55 (91)

www.info.hotims.com/45679-91Kobe Steel Ltd ................................................................48 (82)

www.info.hotims.com/45679-82Linde Engineering North America Inc. ..............................42 (73)

www.info.hotims.com/45679-73Linde Process Plants ....................................................... 14 (85)

www.info.hotims.com/45679-85Paharpur Cooling Towers, Ltd. ......................................... 28 (102)

www.info.hotims.com/45679-102Paqell ......................................................................... 108 (174)


PARCOL SpA ....................................................................15 (152)www.info.hotims.com/45679-152

Pentair Porous Media .....................................................60 (64)www.info.hotims.com/45679-64

Petro-Canada Lubricants ..................................................2 (76)www.info.hotims.com/45679-76

Prosernat ......................................................................45 (163)www.info.hotims.com/45679-163

Quest Integrity Group LLC.................................................31 (157)www.info.hotims.com/45679-157

Samson GmbH ...............................................................66 (167)www.info.hotims.com/45679-167

Sandvik Materials Technology ......................................... 74 (61)www.info.hotims.com/45679-61

Servomex Ltd............................................................... 106 (175)www.info.hotims.com/45679-175

Smith & Burgess LLC ....................................................... 83 (72)www.info.hotims.com/45679-72

SO.CA.P. s.r.l. ................................................................. 41 (161)www.info.hotims.com/45679-161

Spraying Systems Co ...................................................... 10 (66)www.info.hotims.com/45679-66

Summit Industrial Products, Inc. ..................................... 36 (159)www.info.hotims.com/45679-159

T.F. Hudgins, Inc .............................................................65 (166)www.info.hotims.com/45679-166

Team Industrial Services ................................................. 23 (95)www.info.hotims.com/45679-95

Tiger Tower Services ....................................................... 18 (75)www.info.hotims.com/45679-75

Total Safety ...................................................................56 (99)www.info.hotims.com/45679-99

Trachte USA ..................................................................107 (173)www.info.hotims.com/45679-173

UOP, A Honeywell Company ............................................24 Vega Americas, Inc. .........................................................12 (151)

www.info.hotims.com/45679-151Weir Minerals Lewis Pumps ............................................. 16 (94)

www.info.hotims.com/45679-94Zyme-Flow, Decon Technology ........................................89 (92)


This Index and procedure for securing additional information is provided as a service to Hydrocarbon Processing advertisers and a convenience to our readers. Gulf Publishing Company is not responsible for omissions or errors.

Bret Ronk, PublisherPhone: +1 (713) 529-4301Fax: +1 (713) 520-4433E-mail: [email protected]

SALES OFFICES—NORTH AMERICA

IL, LA, MO, OK, TXJosh MayerPhone: +1 (972) 816-6745, Fax: +1 (972) 767-4442E-mail: [email protected]

AK, AL, AR, AZ, CA, CO, FL, GA, HI, IA, ID, IN,

KS, KY, MI, MN, MS, MT, ND, NE, NM, NV, OR,

SD, TN, TX, UT, WA, WI, WY,

WESTERN CANADA Diana Smith Phone/Fax: +1 (713) 520-4449Mobile: +1 (713) 670-6138E-mail: [email protected]

CT, DC, DE, MA, MD, ME, NC, NH, NJ, NY, OH,

PA, RI, SC, VA, VT, WV,

EASTERN CANADAMerrie LynchPhone: +1 (617) 357-8190, Fax: +1 (617) 357-8194Mobile: +1 (617) 594-4943E-mail: [email protected]

CLASSIFIED SALES

Gerry MayerPhone: +1 (972) 816-3534, Fax: +1 (972) 767-4442E-mail: [email protected]

DATA PRODUCTS

Lee NicholsPhone: +1 (713) 525-4626, Fax: +1 (713) 520-4433E-mail: [email protected]

SALES OFFICES—EUROPE

FRANCE, GREECE, NORTH AFRICA, MIDDLE EAST,

SPAIN, PORTUGAL, SOUTHERN

BELGIUM, LUXEMBOURG, SWITZERLAND,

GERMANY, AUSTRIA, TURKEYCatherine WatkinsTél.: +33 (0)1 30 47 92 51Fax: +33 (0)1 30 47 92 40E-mail: [email protected]

ITALY, EASTERN EUROPEFabio PotestáMediapoint & Communications SRLPhone: +39 (010) 570-4948Fax: +39 (010) 553-0088E-mail: [email protected]

RUSSIA/FSULilia FedotovaAnik International & Co. Ltd.Phone: +7 (495) 628-10-333E-mail: [email protected]

UNITED KINGDOM/SCANDINAVIA,

NORTHERN BELGIUM, THE NETHERLANDSMichael BrownPhone: +44 161 440 0854Mobile: +44 79866 34646E-mail: [email protected]

SALES OFFICES—OTHER AREAS

AUSTRALIA—PerthBrian ArnoldPhone: +61 (8) 9332-9839Fax: +61 (8) 9313-6442E-mail: [email protected]

CHINA—Hong KongIris YuenPhone: +86 13802701367, (China) Phone: +852 69185500, (Hong Kong)E-mail: [email protected]

BRAZIL—São PauloAlfred BilykPhone/Fax: 11 23 37 42 40Mobile: 11 85 86 52 59 E-mail: [email protected]

INDIAManav KanwarPhone: +91-22-2837 7070/71/72 Fax: +91-22-2822 2803Mobile: +91-98673 67374E-mail: [email protected]

INDONESIA, MALAYSIA, SINGAPORE,

THAILANDPeggy ThayPublicitas Singapore Pte LtdPhone: +65 6836-2272Fax: +65 6634-5231E-mail: [email protected]

JAPAN—TokyoYoshinori IkedaPacific Business Inc.Phone: +81 (3) 3661-6138Fax: +81 (3) 3661-6139E-mail: [email protected]

KOREAD. S. ChaiDongmyung Communications, Inc.Phone: +82 (2) 391 4254Fax: +82 (2) 391 4255E-mail: [email protected]

PAKISTAN—KarachiS. E. AhmedIntermedia CommunicationsPhone: +92 (21) 663-4795Fax: +92 (21) 663-4795

REPRINTS

Rhonda Brown, Foster Printing ServicePhone: +1 (866) 879-9144 ext. 194E-mail: [email protected]


Engineering Case Histories

A. SOFRONAS, CONSULTING ENGINEER

http://mechanicalengineeringhelp.com

Case 72: Interaction between disciplines when troubleshooting

When troubleshooting, we can become fixated on the dis-cipline that we are most familiar—in the author’s case, me-chanical engineering. Investigating the failure of a machine may have a group of mechanical engineering specialists in-volved in the troubleshooting effort. When one’s specialty is vibration analysis, the first thought for the failure could be a torsional or linear vibration problem. A stress analyst will focus on developing an analytical model to describe this fail-ure. Conversely, the materials engineer will be investigating corrosion issues as the root cause. Likewise, the process and control engineers will be reviewing computer simulations and searching for possible abnormal operations.

It’s human nature to want to analyze problems with the tools that you are most familiar in using. I certainly have done this during my career.

Solution. When the solution is obvious, then the one-disci-pline approach can work. A specialist may have seen a very similar failure and solved the problem. Indeed, most day-to-day failures are resolved by part replacement or with field per-sonnel applying a simple fix. Replacing damaged bearings be-cause the oil was contaminated, and correcting the root cause for the contamination, will mitigate future failures.

Silver-bullet solutions. A silver-bullet, single solution does not work well on complex multivariable system prob-lems. Erroneous solutions can result in severe consequences to equipment, personnel and careers with the one-discipline approach.

This can be addressed by using a structured-team prob-lem-solving approach, with all of the necessary disciplines represented and a team leader who has been selected by management. In this way, all of the potential causes can be identified, using the multi-discipline team. There is no pur-pose in outlining technical problem-solving methods that are well documented.1, 2 Most of us have attended seminars or have been involved in such sessions. We are familiar with the benefits.

Follow through. I have noticed during my career that there are times when the structured-team methods have not been applied correctly. There are many reasons for the lapse in ap-plication and include:

• A deadline is approaching, and it is imperative that this deadline be met. Result: The uninformed make critical deci-sions.

• No one knows that such an approach will help, and no one has been trained for it.

• The problem solving is well under way before it is real-ized that the approach is chaotic, and that a more structured method is needed.

• There is not much urgency in solving the problem; no one wants to spend any money, delegate personnel or be responsible.

• The solution to the problem is started with no action plans and no definite responsibilities.

• Various disciplines that are needed to solve the problem are not available at the site.

Who is responsible? It is management’s job to address all of these listed reasons. The question becomes: When is it neces-sary to utilize a structured-team approach to solve a problem?

From my viewpoint, when safety, litigation concerns, ma-jor production losses or careers may be at risk, then a struc-tured-team approach is needed. Many excellent articles have been written on when just making a repair is inadequate and more should be done.3 A quick computer search on engineer-ing disasters will show many such failures, and, unfortunately, most were avoidable. There are many reasons to consider structured-team problem-solving approaches, and training in root-cause analysis methods, such as the Kepner-Tregoe or TapRooT methods.

Due to the present litigious society, it is very prudent to invest in methods that produce a “paper trail” and show that a sincere and thorough effort was made to address the concern. Detailed and concise calculations provide a method to docu-ment that a sincere effort has been made to understand the concern. Other ways are to include the input of well-known technical experts familiar with the equipment or process in-volved and have experience with similar failures.

LITERATURE CITED 1 Bloch, H. P. and F. K. Geitner, Machinery Failure Analysis and Troubleshooting, Gulf

Publishing Company, Houston, p. 343. 2 Kepner, C. H., and B. B. Tregoe, The Rational Manager: A Systematic Approach to

Problem Solving and Decision Making, Kepner-Tregoe, Second Ed., 1976. 3 Bloch, K., “Extreme failure analysis: Never again a repeat failure,” Hydrocarbon

Processing, April 2009, pp. 87–97.

DR. TONY SOFRONAS, P.E., was worldwide lead mechanical engineer for ExxonMobil Chemicals before retiring. He now owns Engineered Products, which provides consulting and engineering seminars on machinery and pressure vessels. Dr. Sofronas has authored two engineering books and numerous technical articles on analytical methods. Early in his career, he worked for General Electric and Bendix, and has extensive knowledge of design and failure analysis for various types of equipment.

Customer: Oil refinery, Russia.

Challenge: High reliability in extreme operating conditions.

Result: Elliott centrifugal compressors have operated for more than twenty years without overhaul.

They turned to Elliottfor reliable compression solutions.

Many of Russia’s largest refineries turn to Elliott compressors and turbines for critical applications such as hydrotreating and hydrocracking, fluid catalytic cracking (FCC), coking and alkylation. Who will you turn to?

C O M P R E S S O R S T U R B I N E S G L O B A L S E R V I C E

EBARA CORPORATION

www.elliott-turbo.com

The world turns to Elliott.


Stimulate the heart of your hydroprocessing unit

www.axens.net

ImpulseTM, the catalyst technology that combines the stability you recognize with the activity you need


Processing May 2013

Documents

Transcript of Processing May 2013