gas cov copy 5 - DigitalRefining | Refining, Gas and ... your raw natural gas contain hydrogen...

gas 2013

PTQ supplement

gas cov copy 5.indd 1 13/03/2013 12:24

1209

_e

www.airliquide.com

Whatever the impurity, whatever the composition, Air Liquide Global E&C Solutionshas the right treatment.

Does your raw natural gas contain hydrogen sulfi de,carbon dioxide, mercaptans or more?

The composition of natural gas varies tremendously: almost every source contains a diff erent blend of impurities. The options for treatment are almostas diverse. That’s why off ering a solution specifi cally designed for your gas fi eld

is crucial. We as your partner of choice provide solutions for all types of natural gas, including associated and unconventional gas, from a single source. Customised and effi cient.

air liquide.indd 1 12/03/2013 14:21

http://www.eptq.com/jump.aspx?a=Air&p=gas13

©2013. The entire content of this publication is protected by copyright full details of which are available from the publishers. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means – electronic, mechanical, photocopying, recording or otherwise – without the prior permission of the copyright owner.The opinions and views expressed by the authors in this publication are not necessarily those of the editor or publisher and while every care has been taken in the preparation of all material included in Petroleum Technology Quarterly the publisher cannot be held responsible for any statements, opinions or views or for any inaccuracies.

3 Shale gas challenges ChrisCunningham

5 Mercury removal processes NeilEckersleyandDavidRadtke UOP, A Honeywell Company LeonRogersandShawnBrennan Enterprise Products 17 Solubility of hydrocarbons and light ends in amines NathanHatcher,ClaytonJonesandRalphWeiland Optimized Gas Treating, Inc

27 Industrial gases support the natural gas production chain StephenHarrisonandErnstMiklos Linde Gases 33 Formulated solvent reduces shale gas processing costs RobertLDotson Dow Oil, Gas & Mining

43 Qatar’s LNG industry MeghnaBahlandSandeepKumar Fluor

47 Flame interaction and rollover solutions in ethylene cracking furnaces RexKIsaacs Zeeco

53 Steam reforming natural gas containing higher hydrocarbons VishwasDeshpandeandSubrataSaha Reliance Ports & Terminals Limited

Gazprom’sSurgutcondensatestabilisationplant. Photo: Gazprom

2013www.eptq.com

gasptqYLRETRAUQYGOLONHCET MUELORTEP

ed com.indd 1 13/03/2013 13:45

enersul.indd 1 6/12/12 13:25:03

Select 68 at www.HydrocarbonProcessing.com/RS

enersul.indd 1 12/03/2013 17:03

http://www.eptq.com/jump.aspx?a=Enersul&p=gas13

The European Union has arguably been the global leader in biodiesel production and use, with overall

biodiesel production increasing from 1.9 million tonnes in 2004 to nearly 10.3 million tonnes in 2007. Biodiesel production in the US has also increased dramatically in the past few years from 2 million gallons in 2000 to approximately 450 million gallons in 2007. According to the National Biodiesel Board, 171 companies own biodiesel manufacturing plants and are actively marketing biodiesel.1. The global biodiesel market is estimated to reach 37 billion gallons by 2016, with an average annual growth rate of 42%. Europe will continue to be the major biodiesel market for the next decade, followed closely by the US market.

Although high energy prices, increasing global demand, drought and other factors are the primary driv-ers for higher food prices, food competitive feedstocks have long been and will continue to be a major concern for the development of biofu-els. To compete, the industry has responded by developing methods to increase process efficiency, utilise or upgrade by-products and operate with lower quality lipids as feedstocks.

Feedstocks

Biodiesel refers to a diesel-equivalent fuel consisting of short-chain alkyl (methyl or ethyl) esters, made by the transesterification of triglycerides, commonly known as vegetable oils or animal fats. The most common form uses methanol, the cheapest alcohol available, to produce methyl esters. The molecules in biodiesel are primar-ily fatty acid methyl esters (FAME), usually created by transesterification between fats and methanol. Currently, biodiesel is produced from various vegetable and plant oils. First-genera-tion food-based feedstocks are straight vegetable oils such as soybean oil and animal fats such as tallow, lard, yellow grease, chicken fat and the by-products of the production of Omega-3 fatty acids from fish oil. Soybean oil and rapeseeds oil are the common source for biodiesel production in the US and Europe in quantities that can produce enough biodiesel to be used in a commercial market with currently applicable technologies.

First-generation feedstocks for

Gas 2013 3

Editor Chris Cunningham [email protected]

Production EditorRachel [email protected]

Graphics EditorRob [email protected]

Editorial tel +44 844 5888 773fax +44 844 5888 667

Business Development DirectorPaul [email protected] Advertising SalesBob [email protected]

Advertising Sales Officetel +44 844 5888 771 fax +44 844 5888 662

PublisherNic [email protected]

CirculationJacki [email protected]

Crambeth Allen Publishing LtdHopesay, Craven Arms SY7 8HD, UKtel +44 844 5888 776fax +44 844 5888 667

ISSN 1362-363X

ptq (petroleum technology quarterly) (ISSN No: 1632-363X, USPS No: 014-781) is published quarterly plus annual Catalysis edition by Crambeth Allen Publishing Ltd and is distributed in the US by SP/Asendia, 17B South Middlesex Avenue, Monroe NJ 08831. Periodicals postage paid at New Brunswick, NJ. Postmaster: send address changes to ptq (petroleum technology quarterly), 17B South Middlesex Avenue, Monroe NJ 08831.Back numbers available from the Publisher

at $30 per copy inc postage.

Shale gas challenges

Shell Canada’s Shantz sulphur-forming plant is a sight to behold; several sights in fact. Its great blocks of primrose yellow byproduct recovered from Alberta’s gas-rich foothills extend for well over 100m, all at neck-craning

height. Before the next train arrives to load up, sulphur is extracted from the blocks for remelting, forming into pellets and storage in giant silos. When loading time arrives, a 2km-long train, pulled and pushed by two giant locomotives, encircles the whole complex. As the train’s 100-plus wagons move into place, a loading hopper drops 100 tonnes of pellets into a wagon in the space of about 30 seconds.

With its 10 000-tonne load, the train moves off in a stately procession across the Rockies to the Vancouver terminal. From there, the sulphur is shipped to chemical fertilizer producers in China to transform the productiv-ity of local farms.

All of the above is a lead-in to say that these sights are less likely to be seen in the vicinity of a shale gas play. Old sulphur hands in Calgary happily report how drillers have stumbled across conventional gas wells with a hydrogen sulphide content greater than 90%, but there is, relatively speaking, very little of that toxic gas in the raw output squeezed from shale measures.

As momentum builds across the globe towards dominance by shale gas operations among new production sites, gas treating engineers have a revised set of challenges to deal with in their approach to plant design and operation.

In this issue of PTQ’s Gas supplement, Dow’s Robert Dotson explains the challenges of treating raw gas from shale plays and highlights a problem caused by too low concentrations of H2S on the amine plant when shale gas is treated in conventional amine formulations. A combination of DEA with the other main component of gas acidity, carbon dioxide, gives rise to thermal degradation of plant, a problem countered by the formation of a protective iron sulphide coating when H2S is present.

While MDEA will enable higher amine concentrations to be employed without the problems arising from CO2 degradation products, it is more limited than MEA and DEA for the primary purpose of CO2 removal.

Dotson explains how a replacement solvent, specially formulated for shale gas operations, transformed the operating costs of a leading gas processor.

Previously in PTQ Gas, Ralph Weiland and Nate Hatcher of Optimized Gas Treating have explained the challenges of treating shale gases with very low H2S-to-CO2 ratios in order to meet pipeline specifications for CO2 content. The essential challenge is that only a small flow of solvent is needed to treat a large volume of gas, with maximum selectivity for H2S. The H2S content needs to be reduced to pipeline specification without removing too much CO2 at the same time.

Weiland and Hatcher’s approach centres on tray designs specifically tailored to shale gas treating, where there is a need for maximum absorption of H2S while keeping CO2 removal to the minimum level possible. Their premise is that trays exhibit very different mass transfer characteristics when they are operating hydraulically in the froth and spray regimes, and that the mass and heat transfer characteristics of trays need to be modelled precisely to meet the requirements of H2S removal and the necessary level of CO2 preservation.

CHRIS CUNNINGHAM

ptqYLRETRAUQYGOLONHCET MUELORTEP

ed com.indd 2 13/03/2013 11:07

cbi.indd 1 12/03/2013 14:27

http://www.digitalrefining.com/64,sponsors,CB_I_.html#.UVAlZ46kJ95

Mercury removal processes

Mercury is present in many of the world’s natural gas fi elds. Process plants with

brazed aluminium heat exchangers, including LNG facilities and nitro-gen rejection units, are particularly susceptible to corrosive attack by mercury. There is an increased awareness on the part of gas processors to better protect their assets and address environmental concerns by removing mercury at the most appropriate location from their facilities

The use of sulphur-impregnated activated carbon has been prevalent in protecting process plant equip-ment from mercury ingress via natural gas streams. However, over the last decade, carbon-based options have been replaced in many facilities by the use of more specifi c base and noble metal- promoted, non-regenerative and regenerative solutions.

A comparison of these more up-to-date mercury removal process solutions is made in three case studies, and the individual plant drivers leading to the require-ment to remove mercury are discussed.

Why remove mercury?Mercury is a naturally occurring element found in small but measur-able concentrations in an increasing number of hydrocarbons globally. From refi neries to natural gas plants, from coal-fi red power stations to petrochemical produc-tion facilities, mercury is becoming more prevalent and problematic, and technologies to mitigate against the effects of mercury are in demand more than ever before.

A careful evaluation of the options for removing mercury from natural gas plant feed and product streams is a prudent exercise

NEIL ECKERSLEY and DAVID RADTKE UOP, a Honeywell Company LEON ROGERS and SHAWN BRENNAN Enterprise Products

Mercury is often associated with natural gas, condensates, C3-C6 refi nery product streams (such as naphtha, gasoline and LPG) and petrochemical feed streams. Each of these hydrocarbons is challenged in various ways when mercury is present, and process plants and pipeline assets often require a complete removal of mercury as a result.

Removal and capture of mercury is important for a number of reasons: • Process plants with brazed aluminium heat exchangers are susceptible to corrosive attack by mercury, and alloys of aluminium are prone to liquid metal embrittle-ment (LME), causing serious structural damage, particularly when liquid mercury comes into contact with air or water • Product streams such as conden-sates, LPG, NGL and naphtha (in

the case of refi neries) are less valu-able when “distressed” by mercury• Many refi nery and petrochemical catalysts are poisoned by mercury. Mercury has the ability to manifest itself in many crude column prod-ucts and is measurably present in many downstream unit operations. Since increasingly refi ners are set up to sell downstream crude frac-tions to analogous petrochemical customers, consideration should be given in removing mercury even from trace levels to gain more value from individual petrochemical feed streams • Mercury may have health and safety impacts in certain applications.

Types of mercury associated withhydrocarbon streams Mercury takes on several different chemical forms, depending on the hydrocarbon in question. Figure 1

www.eptq.com Gas 2013 5

Natural gas Condensate

Crude oil

Trace suspended Some organicSome suspended

ElementalPrevalent

Some ionicTrace organic

Trace suspended

Trace ionic

Type of mercuryElementalIonicOrganicSuspended

Figure 1 Forms of mercury in common process plant hydrocarbon streams

uop.indd 1 13/03/2013 10:55

6 Gas 2013 www.eptq.com

LME can cause crack initiation and propagation, particularly in the proximity of a weld.2 In order to safeguard against the catastrophic failure of cryogenic equipment, typical maximum levels of mercury are now required in and around these valuable cold boxes within gas processing trains. One level that has found prominence is that the gas entering cryogenic equip-ment contains no more than 10 ngHg/Nm3 gas.

Importance of measuring mercury Mercury needs to be measured in order to determine which mercury removal option will provide the most cost-effective solution to meet desired results. Whether it is simply removing the mercury from a process stream to meet specifi ca-tion, protecting the entire plant or ensuring mercury removal for envi-ronmental compliance, mercury levels must be known. In order to properly design a mercury removal system, accurate mercury measure-ment is critical to properly size the system and to avoid having a system that is overly large and uneconomical or too small to satisfy the required outlet mercury

lists four forms of mercury. These discrete categories exist in natural gas, condensate and crude oil.

Elemental and organic mercury fall into the category of being hydrocarbon soluble. Ionic mercury species are water soluble and comprise examples that include both sulphate and chloride salts of mercury (HgSO4 and HgCl2). Suspended mercury is a broadly defi ned descriptor comprising particulates including mercury-containing species such as HgS.

Levels of mercury in natural gasDifferent areas of the world have varying levels of mercury in their natural gas reservoirs. Figure 2 shows average mercury levels that have been reported to UOP. In recent years, mercury levels have increased from typical highs of 30 or 40 ug/Nm3 to levels exceeding 1000 ug/Nm3 in the Pacifi c Rim area. With a greater understanding of levels in specifi c geographical areas has come a greater level of expectation in terms of what is required to remove mercury both on- and off-shore in a variety of locations worldwide.

Need to protect cryogenic equipmentA well-known reason to remove mercury in a natural gas processing plant is to protect brazed alumin-ium heat exchangers and the cold box in nitrogen rejection units, to prevent the compromising of these valuable pieces of equipment. In the early 1970s, trace levels of mercury accumulating in the cryo-genic recovery section of an LNG production plant at Skikda, Algeria,1 caused catastrophic failure of a heat exchanger. It was found that a combination of mercury and water at temperatures around 0°C caused corrosion in aluminium tubes constructed from aluminium alloy 6061. Subsequent studies revealed far more data on the mechanistic details of how mercury reacts with aluminium, with aluminium diffusing into the mercury droplet followed by conversion to Al2O3 by reaction with air and water.

The consequence is that mercury bores into aluminium and signifi cantly compromises aluminium-containing equipment. Specifi cally, LME has been respon-sible for a number of failures in the 40 years since the Skikda incident.

1-100µg/Nm31-100µg/Nm3The Americas

1-130µg/Nm3North Africa

1-500µg/Nm3Europe

50-200µg/Nm3Australia

200-2000µg/Nm3Asia-Pacific

Gulf

Off-shore

Figure 2 Mercury levels reported in natural gas reservoirs in various geographic locations

uop.indd 2 13/03/2013 10:55

specifications. For existing units, mercury levels in the feed must be monitored for changing inlet levels that might exceed the designed capabilities of the mercury removal unit (MRU). Finally, monitoring the mercury levels exiting the mercury removal unit is critical to verifying proper performance and protection of down-stream equipment.

Where the sampling and testing is conducted is also critical to the collection of mercury data that will be valuable in the evaluation of the plant. It is not just a matter of selecting the most convenient sample point. Configuration of piping and dead legs, or stagnant flow sections of piping, must be taken into considera-tion when selecting locations for sampling. Not all plants and processes are alike; understanding the differences will help ensure that the correct data are gathered from locations that will provide valuable information for the challenge at hand. Overall process knowledge is crucial, as this helps not only with know-ing where to look for mercury but how the process could affect the distribution of mercury throughout the system.

Once the appropriate sampling locations have been identified, the most important task in mercury analysis begins: sampling. It is critical to have as little process piping (or tubing) as possible between the process pipe containing the flowing process stream to be tested and the sample collection system. In fact, zero would be ideal. As much metal as possible must be eliminated from the sampling path in order to minimise interfer-ences that can be caused by mercury adhering to metal surfaces and then being transportable in unpredictable ways that can result in false high mercury measure-ments. If any part of the sampling system must be metal, the metal should be heated. This will minimise or eliminate the ability of the mercury to stick to the metal portions of the sampling system and drastically reduce measurement errors caused by this phenome-non of mercury. Current gas-phase analysis methods generally require that the pressure during testing be reduced to essentially atmospheric pressure. As soon as the pressure is reduced, all efforts should be made to switch the sample system components to Teflon, which has proven to be very resistant to the hold-up of mercury. The goal is to eliminate the possibility of transient mercury depositing in the sampling system, which could move during sample collection and cause inaccurate mercury readings. In addition, it is crucial that the sample point be purged using a continuous flow of process fluid. This purging establishes a steady state in the sample point, allowing the greatest proba-bility of obtaining a sample that is truly representative of the process stream being tested. It is also recom-mended to heat the trap used to collect the mercury sample in order to reduce any chance of interference, such as hydrocarbons blinding the surface of the trap and reducing the efficiency of mercury capture.

Once the sample is collected, it must be analysed for mercury. There are a number of mercury analysers available using various analytical techniques. The two most common analytical techniques for mercury


analysis are cold vapour atomic fluorescence spectros-copy (CVAFS) and atomic absorption with Zeeman correction. The analyser most commonly used by UOP is the Tekran Model 2600. It is favoured due to its high sensitivity, which reduces sample size requirements, and, in turn, allows for quicker sample collection and more frequent sampling. With appropriate sampling techniques, both methods can be used for analysis of gas or liquid samples. Many have found the CVAFS method to be the most reliable, with no interference issues encountered to date. The CVAFS method includes dual-stage traps, which allows for two purge steps and ensures contaminants that may affect accu-racy are eliminated. Operation of the various mercury analysers is straightforward and the results of the sample are generally available within 15 minutes after the sample is taken.

Concerns around mercury in natural gasMercury in natural gas and natural gas liquid is likely to be in the form of elemental mercury. Although mercury has a high boiling point (357°C), it also has a high vapour pressure, which means that it is very mobile. This leads to a specific issue for today’s gas processor, in that it can disperse throughout gas plant assets and present issues in terms of how and where it should be removed. Left unchecked, mercury will deposit on surfaces, including those common to

president truly meant what he said and his words were more than political rhetoric, we could see an end to the anti-fossil fuel environment that has persisted throughout his first term in office. I am hopeful that in 2013 this administration will work to advance a true “all-of-the-above” energy strategy that recognises the importance of all of our domestic energy resources and fuel and petrochemical manufacturers in rebuilding our nation’s economy.

The US, in combination with our ally and friend Canada, has an abundance of natural resources capable of bringing the nation to economic prosperity and North American energy independence within the next 10 years. This can be done by expanding shale development on federal lands, stopping the attacks on hydraulic fracturing that threaten affordable feedstocks necessary for all manufacturing, and increasing offshore drilling permits.

Further, by approving the Keystone XL Pipeline, the president could send a strong signal to the nation that he is serious about creating thousands of domestic jobs and improving our economy. More importantly, the president would put the world on notice that the US is charting a future to include energy independence and security for the country, with a goal of ultimately becoming a global energy provider.

Sadly, without a significant practical and political course correction, growth is not possible for fuel and petrochemical manufacturers in the current regulatory environment. In January 2011, President Obama ordered a review of federal regulations to be eliminated because they hinder economic growth and job creation. Two years later we are still waiting. A commitment by the administration is needed to fix what continues to be a regulatory nightmare for the refining and petrochemical industry and an unnecessary, costly burden for the American public. For example, federal regulations require fuel and petrochemical manufacturers to spend billions of dollars to reduce greenhouse gas (GHG) emissions, even though the EPA has acknowledged the reductions would bring little or no environmental benefit. These same regulations increase energy costs, result in job loss, and harm the US economic and national security.

AFPM will continue to support sensible and beneficial environmental regulation, but we believe that America’s national interest would best be served by comprehensive and objective cost-benefit analyses of regulations to determine which make sense and which do more harm than good.

Today, it is clear that the vast majority of Americans want to develop our own natural resources and promote manufacturing jobs. Our nation is blessed with an abundance of energy resources that could revitalise job growth and our economy, enhance our national security, and ensure a strong fuel and petrochemical manufacturing industry. The decisions made by the Obama administration in 2013 will determine the future of the industry I represent and, more importantly, the entire nation.

Dmitry BalandinChief Financial OfficerGazprom Neftekhim Salavat

We are living through difficult economic times. Despite this, the market

for oil and gas production in Russia remains stable. The main source of this stability is high oil prices, which allow OJSC “Gazprom Neftekhim Salavat” and other major market participants to achieve positive margins. The domestic market still tends to be influenced by the activities of government and the regulatory authorities, in particular the measures government takes to reduce domestic fuel prices. During the second half of 2011, the government implemented a new export duty system (so-called “60-66”), which benefits the upstream industry, but is less advantageous for companies within the downstream sector. Thanks to our broad market appeal and high-quality offerings, Gazprom Neftekhim Salavat is finding the export and domestic markets to be profitable.

There are challenges that Russian process industry companies operating in the oil and gas market need to overcome in the next 12 months. One key challenge is the high volatility of oil prices and crack spreads for oil

www.eptq.com PTQ Q1 2013 11

nitrogen removal. The extent of FCC yield improve-ments were often a function of desired operational cycle life and available hydrogen for the pretreat units. Hydroprocessing catalyst systems were developed util-ising cobalt molybdenum (CoMo) and nickel molybdenum (NiMo), depending on these objectives and constraints.

In today’s clean fuel operations, much investment has been made in ULSD and FCC naphtha HDS, with few refiners now achieving environmental compliance via previously designed pretreat units. Additional results of the global drive towards clean fuels are continued advances in catalyst technology that have provided significant gains in both HDS and HDN performance. These technology gains are being utilised to drive new FCC pretreat designs to very high levels of performance and have provided refiners with the option of revisiting how best to maximise the value of existing FCC pretreat units. This has resulted in many units shifting catalyst system designs in order to provide higher levels of nitrogen removal and aromatic saturation by using more high-activity NiMo catalysts, maximising FCC conversion capability.

If distillate maximisation is desired, many FCC pretreat units can be revamped to effectively operate in a MHC mode. This more severe operation is performed with higher reactor temperatures and often by modify-ing the catalyst system to include a more active

conversion catalyst such as an amorphous silica-alumina (ASA) or zeolite.

Depending on the conversion and distillate selectiv-ity required, all alumina, alumina/ASA or alumina/ zeolite stacked systems can be considered. Higher conversions can be achieved by using alumina/ASA stacks and even higher by using alumina/zeolite stacks compared to a total alumina system. In specifying a MHC catalyst system, the balance of hydrotreating versus cracking catalyst and the potential addition of reactor volume is largely influenced by feed qualities and the desired level of conversion. As many of the feeds processed are high in contaminant metals, sulphur and nitrogen, the pretreat section is required to remove these contaminates to ensure a sufficient cycle life can be maintained while both meeting any product targets and minimising nitrogen slip into the cracking section of the reactor. Feed quality and the reactor and catalyst system specified determine the ultimate sulphur and nitrogen removal capability for a given cycle life. HDS functionality can remain an important criteria for some MHC units depending on existing product specifications, which are dependent on site refinery constraints and capabilities. However, HDN capability is often more important, as it influ-ences cracking catalyst selection and performance due to the remaining nitrogen heteroatoms, which reduce cracking reactions. As mentioned, zeolite-containing products can provide the highest levels of conversion, but they tend to be the most sensitive to nitrogen slip, reducing their long-term effectiveness in such cases. Amorphous silica alumina (ASA) cracking catalysts provide increased levels of nitrogen tolerance with a lower level of conversion capability and, for units with limited HDN capability, conventional pretreat catalyst can be operated in a MHC mode but with a reduced conversion capability.

When using desalter water for coke cutting, you should at least consider some major problems related to water composition: entrained salts; entrained caus-tic; and entrained sludge and sediments. Entrained inorganics might have an impact on the metallurgies of the cutting tool, coke drum and lines, together with potential salts precipitation in the cutting tool’s nozzles. Some mitigation of such phenomena might be found in acidification and/or anti-scalant injection, but all of that needs to be carefully evaluated.

Also consider that caustic, oil carry-under and chem-icals might act as emulsifiers and can potentially stabilise the frothing of coke particles, preventing/limiting their precipitation in the water recovery system and creating coke particles carry-over in the cutting tool, which, in turns, creates plugging and possible erosion/corrosion.

MINIVAP ON-LINE

Phone +43 1 282 16 27-0 | Fax +43 1 280 73 [email protected] | www.grabner-instruments.com

VAPOR PRESSURE PROCESS ANALYZER

VP of Gasoline, Crude Oil and LPG•According to Standards

RVP, DVP, TVP and V/L = 20•

Highest Precision •

Up to 2 Sample Streams•

Adjustable Temperature•Range: 20-60°C

Adjustable V/L Ratio•

Automatic Calibration•

Fast and Easy Maintenance•

Automatic Lubrication•

Q&A copy 12.indd 2 11/9/12 13:23:03outlook copy.indd 5 10/12/12 13:05:23uop.indd 3 13/03/2013 13:38

http://www.eptq.com/jump.aspx?a=grabner&p=gas13


point and entrained liquids are common. This is where fixed-bed, metallic-based MRU products find greatest success in their ability to treat “wet” gas streams in up-front positions.

Non-regenerative metal sulphides can successfully remove mercury from raw gas, upstream of the amine unit and the dehydration vessels. Utilising larger MRU vessels, this approach protects the brazed aluminium heat exchanger and ensures significantly less mercury contamination in and around the process plant. This option has become increasingly popular, since it minimises the total mercury present before there is any opportunity for mercury to migrate to various locations within a gas processing plant and avoids the risk of subsequent partitioning into processed natural gas and conden-sate streams.4 It also avoids subsequent adsorption onto any pipeline asset or piece of equip-ment downstream.

Liquids carry-over onto sulphur-impregnated activated carbonThe degree of liquid adsorbed onto carbon during the life of a sulphur-impregnated carbon-based MRU is shown in Table 1. The MRU can be shown in terms of total wt% volatiles (200°C). Layer 1 represents carbon recovered from the inlet portion of the bed, and layers 2, 3 and 4 represent subse-quent layers. Each layer was bagged upon discharge, following a lifetime measured in months rather than years, as per design.

The total wt% of active sulphur was measured on each layer of discharged carbon. While sulphur levels on newly installed activated carbon are typically 10-18 wt%, the levels on the spent material were measured at 6-9 wt%. In order to measure the mercury removal effi-ciency of sulphur-impregnated activated carbon, the percentage of utilised sulphur was also measured on a molar basis. The percentage of sulphur utilised in the equilibrium section of the vessel (layers 1, 2 and 3) was measured at ~4 wt%. By comparison, on a dry natural gas without the attendant issues of

regenerative adsorbent solutions for mercury contaminant removal.

Regenerative adsorbents formercury removalThe protection of aluminium heat exchangers can be accomplished using a layer of silver-containing molecular sieve inside the dehydra-tion vessels. The active silver forms an amalgam with the mercury, and its zeolitic substrate adsorbs mois-ture in the gas to be treated. This article later examines a case study at Enterprise Meeker, describing this system in operation. This approach offers flexibility in being regenerable, as the mercury- containing gas is bypassed around any cryogenic equipment into the sales gas. If necessary, condensed mercury can be collected and the mercury-entrained gas further treated with a small, non-regenera-tive guard bed, so that it is not passed to the sales gas.

Non-regenerative adsorbents formercury removalThere are two types of non- regenerative MRU: carbon systems and metal sulphide beds. The common and traditional approach to mercury removal has historically been through the use of sulphur-impregnated carbon beds. Existing sulphur-impregnated acti-vated carbon options tend to be less effective at positions upstream of molecular sieve drying systems or glycol injection due to the risk of capillary condensation in the micro-pores of the carbon sub-structure. The pore size distribution of carbon products is such that this has been problematic in the past, particularly where MRU locations have been in the up-front position, where raw gas is often at or close to its dew

pipelines and plant assets. Mercury can then desorb back into gas streams, passing through contami-nated pipelines. As a result of this, prolonged periods of time can elapse between the installation of an upstream MRU and the complete purging of a pipeline.3 This creates a difficult situation for a gas processor and one that requires a lot of thought as to how best to remove mercury, often as close to the front end of a natural gas processing or gas transmission system as is practicable.

The issue of mercury ingress into natural gas is not confined to the gas phase either. Natural gas liquids and condensates are also prone to mercury as a contaminant, and systems capable of both gas- and liquid-phase mercury removal are required by today’s gas processor.

Apart from the basic requirement to ensure that adequate mercury removal can be achieved on gases at or close to their dew points, it is important to ensure that MRU reac-tor volumes are minimised where possible, to reduce capital cost and minimise plot space. This is particularly important when posi-tioning an MRU offshore, where spatial constraints are critical. The contact times required by sulphur-impregnated carbon prod-ucts often lead to a larger than practical MRU footprint. The drive to reduce capital budgets and avoid large volumes of spent material has led to gas processors examining technologies other than sulphur- impregnated carbon.

Mercury removal process optionsThe global market has a number of approaches for mercury removal. These options can be categorised as regenerative adsorbents and non-

Bed Total volatiles Total S, wt% Total Hg, wt% Hg/S on Hg/S on a % Sulphurposition (200°C), wt% (dry basis) (dry basis) w/w basis molar basis, % utilisation

Layer 1 21.1 7.02 1.89 0.27 0.0429 4.29Layer 2 21.0 6.57 1.71 0.26 0.0414 4.14Layer 3 26.8 8.32 2.12 0.25 0.0406 4.06Layer 4 24.4 7.97 0.29 0.04 0.0057 0.57

Impact of liquids on mercury removal using activated carbon in the gas phase

Table 1

uop.indd 4 13/03/2013 10:55

Marine Power Plants Turbomachinery PrimeServ

We build rock solid components for a broad range of applications in re neries and petro-

chemicals. Whether for hydrogen production and recovery, desulfurization, uid catalytic

cracking (FCC), PTA, Fertilizer or IGCC – our axial, screw and centrifugal compressors, as

well as our steam turbines and reactors are built to achieve the highest level of availability.

Rede ning reliability today and tomorrow. Find out more at www.mandieselturbo.com

MAN InsideReliable solutions for the long run

Isothermal andGear-type compressors

Centrifugal compressors

Axial compressors

Expanders

Steam turbines

Process-gas screw compressors

Reactors

Typ 02 - MAN Inside - Plant - 297x210.indd 1 2012-10-15 14:39:55manturbo.indd 1 12/03/2013 14:57

http://www.eptq.com/jump.aspx?a=MAN&p=gas13

Reliable SwiSS Quality

API 618Rod load up to 1'500 kN/335'000 IbsPower up to 31'000 kW/42'100 HP

Full Range:

You get moRe than Just aPRocess gas comPRessoR

Lubricated up to 1'000 bar, non-lubricated up to 300 bar

For highest availability: We recom-mend our own designed, in-house engineered compressor valves and key compressor components

Designed for easy maintenance

We are the competent partner with the full range of services – worldwide

→ www.recip.com/api618

Reliable SwiSS Quality

API 618Rod load up to 1'500 kN/335'000 IbsPower up to 31'000 kW/42'100 HP

Full Range:

You get moRe than Just aPRocess gas comPRessoR

Lubricated up to 1'000 bar, non-

For highest availability: We recom-mend our own designed, in-house engineered compressor valves and key compressor components

Designed for easy maintenance

We are the competent partner with the full range of services – worldwide

YouR BeneFit:lowest liFe cYcle costs

burckhardt.indd 1 7/3/12 21:39:05

http://www.eptq.com/jump.aspx?a=Burckhardt&p=gas13


guard bed inlet screen • September 2009: replace mercury guard bed adsorbent (sulphur- impregnated carbon) • December 2010: install stainless steel top-side/bottom-side reboil-ers, repair demethaniser trays, and inlet dehydrator mol sieve replacement• June 2011: replace mercury guard bed adsorbent.

Initial experiences post start-upShortly after plant start-up, operat-ing issues were observed with the inlet screen in the mercury guard bed. Investigations revealed incor-rect design of the inlet screen, which caused abnormal flow patterns and adsorbent attrition.

gas plant located in Opal, Wyoming, is a single-train plant that processes up to 750 MMscfd of natural gas. Gas is supplied to the plant from the Jonah and Pinedale fields. Construction began in 2006 and the plant was started up on 1 February 2008. Plant design includes inlet liquids separation, gas dehydration and mercury removal (see Figures 3 and 4).

As part of the initial plant design, sulphur impregnated carbon was chosen as a mercury guard bed adsorbent.

Since start-up, Enterprise Pioneer has had four turnarounds. The sequence of events, observations and corrective measures is:• February 2009: repair mercury

liquid entrainment, this percentage utilisation would be expected to be >>10 wt%. The data confirm that sulphur-impregnated activated carbon is prone to sulphur dissolu-tion and micropore blocking when treating wet gas.

The data also strongly suggest that the carbon has co-adsorbed a significant quantity of liquid (20-30%) from the raw natural gas. This contributed to its shortened service life.

Reclaiming mercury from spentadsorbentsActivated carbonAfter the carbon is discharged from an MRU, it is usually sent to a specialised plant, where mercury is reclaimed via vacuum distillation. There is no useful purpose for the remaining carbon and it undergoes high-temperature incineration.

Metal sulphides Specialised processes are used for mercury reclaiming from metal sulphides. For both carbon-based and metal sulphide-based adsor-bents, controls are in place to ship material internationally from source (gas processing plant) to destina-tion (reclaim facility). The paperwork and experience required to accomplish such transportation is complex and requires very care-ful consideration.

Molecular sieve-based adsorbents As the mercury is passed to the regeneration gas in properly regenerated beds, the spent adsor-bent contains no mercury, and therefore mercury reclamation is not an option. Reclamation could be evaluated for the recovery of silver.

Case studiesThe following case studies repre-sent real examples of the use of various forms of mercury removal media outlined above and offer examples of the types of circum-stances different facilities may be required to deal with to address mercury contamination.

Case 1: Enterprise Pioneer The Enterprise Pioneer cryogenic

Raw natural

gas

Inlet scrubber

Inlet dehydrators

Inlet filter

Trim heater

Mercury guard bed

Cryo feed

Dust filter

Figure 3 Enterprise Pioneer dehydration and MRU process

TIC

CompressorExpander

Residue gas compressor

Gas/gas exchanger

Reflux exchanger

Booster

Cryo feed

NGLs

Bottom-side reboiler

Cold separator

Demethaniser

Top-side reboiler

Figure 4 Enterprise Pioneer cryogenic plant

uop.indd 5 13/03/2013 10:56

ExxonMobil — Global Leader in Fuels and Lubes Process TechnologiesExxonMobil Technologies are applied across our corporation and by licensees worldwide in a growing list of process industry applications. Take advantage of our vast experience and ongoing commitment to continuously improve our industry leading technologies.

TECHNOLOGIES :

3225 Gallows Road, Fairfax, VA 22037-0001, USAwww.exxonmobil.com/tsl

+1-703-846-2568 fax [email protected]

Advanced Solutions for the World’s Toughest Energy Challenges

Leadership in Process Technology & Catalysis

developmentdeployment

discovery

2010

2000

1990

1980

1970

1960

1950

1940

SELECTIVE LUBE DEWAXING (MSDW)ExxonMobil’s new generation of MSDW technology continues to lead the industry.

Methanol to Gasoline (MTG)MTG converts methanol to sulfur free gasoline, including first commercial CTL application.

FLUID CATALYTIC CRACKINGFCC unit in 1942 would become the industry standard for production of gasoline.

BenzOUT Converts benzene contained in the refinery’s gasoline blend-stock into high octane alkyl-aromatics by reacting benzene with light olefins.

DISCOVERY OF ZSM-5 ZEOLITEThe discovery of ZSM-5 revolutionized many catalytic processes in refining and petrochemical industry.

pOWERFORMING

HydrofinishingCatalyst

MLDWDewaxing

Discovery of MCM-22

exxon.indd 1 12/03/2013 14:36

http://www.digitalrefining.com/74,sponsors,ExxonMobil_Research_and_Engineering.html#.UVAl_Y6kJ94


in December 2010 revealed the same coating of powder in the top of the tower and liquid redistribu-tors. The layer of powder was thicker than was observed in 2009. Powder samples were analysed at two different labs and the results showed sulphur (86.6% and 89.5%) as the major component of the foulant. Given one sample analysis in 2009, followed by two results in 2010, it was confi rmed that the majority of the sample was sulphur. Signifi cant efforts were then put forth to determine the source of the sulphur.

Discussions with the mercury guard bed adsorbent vendor (sulphur-impregnated carbon) revealed that liquid hydrocarbons leached elemental sulphur from the carbon beds. Cooling this gas led to sulphur crystallisation and thus fouling of the metal surfaces of downstream process equipment. Further analysis of approach temperatures made it obvious that fouling of the refl ux exchanger was increasing, resulting in reduced ethane recoveries.

Efforts were made to determine the source of the hydrocarbon through dewpoint estimation. Simulations revealed that the hydrocarbon dewpoint was ~30°F lower than the actual stream temperature; however, the sulphur fouling was still ongoing.

Efforts were made to fi nd a mercury adsorbent that did not contain elemental sulphur. UOP GB-562 contains copper sulphide,

the new adsorbent bed and screens in place, it was hoped that the foul-ing of the refl ux exchanger would stop.

March 2010 shutdown observationsDuring the March 2010 shutdown, the plant experienced failure of the brazed aluminium top side/bottom side reboilers of the demethaniser. As a result, these were blinded off from the process equipment. Blinding of side reboilers not only resulted in signifi cant ethane recov-ery losses, this also masked the continued loss of heat transfer in the refl ux exchanger (due to the continued fouling from the MRU adsorbent beds). In December 2010, stainless steel reboilers were installed and demethaniser tower modifi cations were made to improve ethane recovery. The refl ux exchanger was not cleaned at this time and this signifi cantly impacted ethane recovery after the December 2010 start-up.

Demethaniser tower inspections

Mercury adsorbent attrition resulted in the generation of dust, causing the failure of several fi lter elements and fouling in the inlet heat exchangers due to fi ne particu-lates. As a corrective measure, the mercury guard bed inlet screen was replaced. However, inlet heat exchangers were not cleaned.

A review of approach tempera-ture trends observed over the three turnarounds can be summarised:• A dramatic increase in the approach temperature can be seen, starting in September 2008, due to the increased adsorbent attrition related to faulty inlet screen design• The fi rst turnaround, in February 2009, stopped accelerated attrition of the mercury guard bed adsor-bent, but the heat transfer of the refl ux exchanger did not improve as the exchanger was not cleaned• During the second turnaround in September 2009, all of the inlet exchangers were cleaned (“back puffed”) to remove particulates. Approach temperatures of only 1-2°F degrees were achieved in the fi rst few days after start-up• However, the approach tempera-ture quickly increased without any particulate fouling of the exchanger between the period of time between the fi rst and second turnarounds (see Figure 5).

September 2009 shutdownobservationsA powder coating was discovered in the top of the tower and on the liquid redistribution trays for beds 1 and 2 during the September 2009 turnaround.

As the fi rst step of investigation, samples of the dust were collected and analysed by mass spectrometer (mass spec), scanning electron microscope (SCEM), and carbon, hydrogen, nitrogen and sulphur (CHNS). The analytical results are shown in Table 2.

Based on analytical results from two of the three samples, initial inference was that a major fraction of the contamination could be due to molecular sieve attrition along with some carbon on the surface. However, there was not a good explanation for the fouling mecha-nism of mole sieve dust. Also, with

60

50

40

30

20

10

28/4

/200

7

14/1

1/20

07

1/6/

2008

18/1

2/20

08

6/7/

2009

22/1

/201

0

10/8

/201

0

26/2

/201

1

14/9

/201

1

1/4/

2012

18/1

0/20

12

6/5/

2013

Appro

ach t

em

pera

ture

, ºF

0

Changes in C2 recovery

3rd TA 4th TA2nd TA1st TA

Hg bed screen

problems

Figure 5 Refl ux exchanger approach temperature

Mass spec SCEM CHNSSodium Sodium Sulphur (75%)Aluminum Aluminum Carbon (8%)Iron Oxygen Calcium SilcaMinor components Calcium Iron Sulphur

Analysis of dust samples from a September 2009 shutdown at

Enterprise Pioneer

Table 2

uop.indd 6 13/03/2013 13:38


mercury breakthrough after two years in service. Table 3 is a summary of the MRU’s process conditions.

After the activated carbon was discharged from the reactors, the MRU was refi lled with UOP GB-562 adsorbent (metal based). The GB-562 parallel fl ow reactors at GSP-5 are shown in Figure 6.

Since the UOP GB-562 was commissioned, the plant has contin-uously recorded effl uent mercury levels below the required specifi ca-tion and maintained the start-of-run pressure drop, which is a priority for PTT. This successful performance has been achieved despite the UOP GB-562 treating a liquid-entrained, water-saturated natural gas.

Figure 7 shows the actual mercury infl uent and effl uent levels at the GSP-5 facility since the start-up of the GB-562 MRU. Despite fl uctuating mercury inlet levels, the GB-562 continues to meet desired effl uent specifi cation.

Case 3: Enterprise USAThe Enterprise Meeker I gas plant, located in Colorado’s Piceance Basin, started up in 2007 with an initial gas processing capacity of 750 MMscfd, incorporating 35 000 b/d of NGL recovery. A phase II expansion in the form of sister plant, Meeker II, started up in 2008, resulting in a doubling of capacity to 1.5 Bscfd gas and 70 000 b/d of produced NGL.

MRU is designed to protect an NGL recovery plant, which incor-porates a cryogenic unit that includes a brazed aluminium heat exchanger. The initial charge of mercury removal adsorbent used from plant start-up was sulphur-impregnated activated carbon. The gas plant’s MRU confi guration consists of two paral-lel reactors each designed to process 265 MMscfd of gas.

The MRU is located upstream of the amine plant and the dryers, and is positioned to treat raw gas as it enters the facility. Historically, the gas entering the MRU has contained some liquid hydrocarbon with a TEG carry-over component. The initial charge of activated carbon experienced a premature

which is insoluble in liquid hydro-carbons. The carbon/sulphur material was replaced by GB-562 in June 2011 and the inlet exchangers were cleaned. Refl ux exchanger approach temperature showed a dramatic improvement, 40-0.5°F. The approach temperature has remained stable since June 2011.

Case 2: PTT Thailand The PTT GSP-5 gas plant located at Map Ta Phut, Rayong, Thailand, was commissioned and started up in 2004. The on-shore facility processes raw gas via a pipeline from off-shore gas fi elds in the Gulf of Thailand. The raw gas entering GSP-5 is conditioned to remove CO2, H2O and Hg, and the total gas fl ow treated is 530 MMscfd. The

Figure 6 GB-562 parallel fl ow reactors at PTT Thailand

70

1101009080

60

130

150140

120

5040302010

0 1 2 3 4 5 6 7 8 9 10 11 12

Merc

ury

concentr

ati

on

, µg/N

m3

Months online

0

MRU inletMRU outlet

Figure 7 Mercury inlet and outlet levels of a GB-562 MRU at PTT

Gas treated Natural gasGas fl ow rate 530 MMscfdOperating pressure 48 kg/cm2

Operating temperature 18°CHg infl uent range 50-200 µg/Nm3

Hg effl uent specifi cation <0.01 µg/Nm3

PTT GSP-5 MRU process conditions

Table 3

Gas treated Natural gasGas fl ow rate to mol sieve vessels 750 MMscfdGB-562 operating temperature 30-40°CGB-562 operating pressure 70 kg/cm2

Raw gas Hg concentration to HgSIV 1 Up to 800 ng/Nm3

Regeneration gas Hg concentration to GB-562 Up to 2000 ng/Nm3

Effl uent Hg concentration from GB-562 <10 ng/Nm3

Enterprise Meeker I and II

Table 4

uop.indd 7 13/03/2013 10:56


4 Edmonds B, Moorwood R A S, Szcepanski R, Mercury partitioning in natural gas, GPA European Chapter Meeting, London, March 1996.

Neil Eckersley is the Integration Leader responsible for integrating all aspects of the Thomas Russell Company into UOP. He holds a BS degree in applied chemistry from Sheffield Hallam University in England.DavidRadtke is a Senior Adsorbents Technical Engineer responsible for customer and sales support and preparation of adsorbent designs for UOP adsorbent products. He holds a BS degree in chemical and petroleum refining engineering from Colorado School of Mines in Golden, Colorado.Leon Rogers is the Plant Manager for Enterprise Products’ Pioneer Cryo Gas Plant in Opal, Wyoming. He holds a BS degree in chemical engineering from Brigham Young University and a MS degree in chemical engineering from Oklahoma State University. Shawn Brennan is the Plant Manager for Enterprise Products’ Meeker Gas Plant located in Rifle, Colorado. He holds a BS degree in chemical engineering from Worcester Polytechnic Institute (WPI) in Worcester, Massachusetts.

pros and cons of these options is a prudent exercise. A favourable Capex solution may not provide favourable Opex for the plant and this could have a severe impact on plant profitability in the long run. This article has presented cases where the effect of feed stream qual-ity, coupled with operating factors, influence the performance of an adsorbent in a gas plant and demon-strate how an incorrect selection of adsorbent can lead to frequent and unwanted shutdowns coupled with a reduced operating efficiency of the unit. A prudent approach is needed for selecting the right mercury removal technique.

With thanks to the co-operation of the PTT Thailand GSP-5 facility and to the Enterprise Meeker facility.

References1 Oil & Gas Journal, 15 Sept 1975, 192. 2 Willhelm M S, Risk analysis for operation of…, AIChE, April 2008, New Orleans.3 Jt. UOP/Equistar paper, Hg removal from cracked…, 2004 Ethylene Producers Conference.

Both Meeker I and Meeker II condition raw gas to remove carbon dioxide using amines. UOP Molsiv UI-94 Adsorbent and Hg-SIV1 molecular sieves are used in the drying vessels to remove water and mercury, respectively, from the raw gas prior to it passing to the cryo-genic system. A non-regenerative mercury guard bed adsorbent (UOP GB-562) is installed to remove mercury from the molecu-lar sieve regeneration stream.

The molecular sieve dryers are configured such that two vessels are in adsorption mode and one vessel in regeneration mode at any given time. Each molecular sieve vessel processes 375 MMscfd of feed gas. Table 4 is a summary of the mercury removal process condi-tions for the molecular sieve and the GB-562 MRU.

ConclusionThere are various treatment options available for mercury removal from natural gas plant feed and product streams. A careful evaluation of

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-Pumpen [email protected] · www.hermetic-pumpen.comHERMETIC-Pump Type CAMTV 52/6+6

uop.indd 8 13/03/2013 10:56

http://www.eptq.com/jump.aspx?a=hermetic-pumpen&p=gas13

RAISING 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:

• Meets API 617 Standards

• Smaller Footprint

• Low Maintenance

• Oil-Free Process Gas

• Optimum Control

• Lower Life Cycle Cost

• Greater Reliability

• Optimized Effi ciency

• Proven Track Record

Cameron’s integrally geared compressors bring57 years of reliability to process gas applications.

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 W

GEARED TOWARD

AD00523CC

CAMER-1132_MidstreamAd_PTQ.indd 1 3/8/13 4:04 PMcameron.indd 1 12/03/2013 14:24

http://www.eptq.com/jump.aspx?a=cameron&p=gas13

Solubility of hydrocarbons and light ends in amines

In processing and treating both wellhead and refinery gases for H2S and CO2 removal, an unfor-

tunate side effect is the loss of hydrocarbons and other light ends gases that are also soluble to vary-ing degrees. Hydrocarbon solubility is never good news, unless the solvent is intentionally being used for dew point control. Any hydro-carbons that dissolve into the solvent in the absorber must reap-pear somewhere else and be discharged from the process. Gas removed from the rich solvent via a flash tank immediately downstream from the absorber (so-called “flash gas make”) is mostly hydrocarbon but with high levels of CO2 and H2S. It may need to be treated further in a low-pressure absorber to remove H2S before it can be used as fuel gas, for example, so its flow rate and composition might need to be known with reasonable accu-racy. Incidentally, gas dissolved in the amine solvent is not the only way hydrocarbons can enter the flash tank. Carry-under of gas entrained (but not dissolved) in the solvent can be a much bigger source than dissolved gas if care is not taken to prevent it.

Once the rich solvent enters the regenerator, its hydrocarbon content is almost completely stripped and it enters the sulphur plant along with the acid gas. Low molecular weight hydrocarbons take up air demand and hydraulic volume in the sulphur plant, lowering sulphur recovery capacity and efficiency. BTEX aromatics are real trouble-makers in sulphur plants and can lead to catalyst deactivation in units with acid gas feeds that are lean in

Modelling procedures predict with considerable certainty what the expected hydrocarbon content of an amine solution in a gas processing plant ought to be

NATHAN HATCHER, CLAYTON JONES and RALPH WEILANDOptimized Gas Treating, Inc

H2S. With richer feeds, and espe-cially oxygen-enriched units, temperature moderation problems can result. In some cases, heavy hydrocarbons in the right boiling range can be steam stripped and trapped in the overhead circuit of the regenerator, leading to foaming.

When conditions are right, hydro-carbons can form a second liquid phase. Foaming then becomes almost inevitable, and it is usually alleviated by the spewing of foamy solvent overhead. Although not a solubility issue, foamed solvent may also carry large amounts of hydrocarbons from the absorber bottom and into the flash tank, where they contribute in a large way to flash gas make rates far in excess of what would be expected from solubility alone.

To estimate solubility in amine treating solutions, a good place to start is with the solubility in water, because water makes up at least 85 mol% of most treating solutions. However, amines are organic mole-cules and have a much higher affinity for hydrocarbons than water does. Thus, hydrocarbon solubility in amine-treating solvents should be expected to be higher than in water, and the solubility should depend not only on the particular amine but also on its strength. One might say that the amine “salts” in hydrocarbons. Dissolved acid gases, on the other hand, cause what is termed “salting-out” of hydrocarbons (and fixed gases). When acid gases dissolve in (and react with) the solvent, they produce copious amounts of ionic components such as protonated amine, sulphide and

bisulphide ions, carbonate and bicarbonate and, in the case of primary and secondary amines, amine carbamates. These are all ionic and result in a high total ionic strength. This causes dissolved hydrocarbon and fixed gas concen-trations to be much lower than would be expected simply on the basis of solubilities in unloaded amines. Unfortunately, unloaded amines are usually what have been used to assess hydrocarbon and fixed gas solubilities in amine-treat-ing solutions. Assessments based solely on unloaded solvent data can be erroneously high, by 50% at high loadings.

This article aims to place hydro-carbon and fixed gas solubilities in amine-treating solutions on a solid, generalised footing. The intended result for practitioners is the abil-ity to predict with considerable certainty just what the expected hydrocarbon content of any amine solution in a properly operating plant ought to be. This is particu-larly important for rich solvents under absorber bottom conditions, because the rich amine is the normal, unavoidable source of hydrocarbons in flash gas make and Claus plant feed streams. And it is precisely here that ion-in-duced salting-out is at its strongest. The importance of being able to make this prediction accu-rately is that it provides a quantitative assessment of the situ-ation expected under normal circumstances in a well-managed facility. If the flash gas make rate is higher, there is probably carry-under from the absorber either as entrained gas or as an


ogrt.indd 1 13/03/2013 11:21


specific amine, with contributions from protonated amine, carbamate, bicarbonate, carbonate and, in prin-ciple at least, on bisulphide and sulphide ions, although no meas-urements appear ever to have been reported of the interactions of HS- and S= with any soluble gas (a worthy subject for research).

Salting-inFirst, it is important to recognise that all salting-in data have been collected for the hydrocarbons in completely unloaded amine solvents. Therefore, if simulator comparisons for the rich amine in a flash tank, for example, are made with literature data for the amine in question, the numbers will almost certainly be very different. The literature data ignore the salting-out effect of the ions gener-ated by loading the solvent with acid gas and, as will be seen, the effect is very substantial indeed.

Mather and co-workers have correlated the salting-in parameter for methane and ethane,3 propane6

and propylene7 in MDEA, DEA, DGA, MDEA and DIPA according to:

In(Si) = kC

a

where k (L . mol-1) is a linear func-tion of temperature: k = a + bt(°C). We have developed similar correla-tions for ethylene, n-butane, 1-butene, n-pentane, benzene, tolu-ene and p-xylene using data provided confidentially by several solvent vendors.

For light hydrocarbons, little data seem to exist for the way in which salting-in parameters depend on temperature. Much of the data are at 313.15K (40°C, 104°F), which is close to many absorber bottom temperatures. In any event, the temperature dependence is weak for propane and propylene so, assuming the correction is temperature inde-pendent in those cases, where there are no data is unlikely to be a serious deficiency.

Figure 1 provides evidence in support of the premise that informa-tion for missing components can be inferred reasonably well by appeal

emulsified second hydrocarbon phase. It points troubleshooting in the right direction, and it also allows a downstream flash gas scrubber to be properly sized.

Modelling and theoryThe basis for determining the solu-bility of any sparingly soluble gas, i, uses temperature-dependent Henry’s Law data for pure water, Hi

o and a Setschenow (1892) salting (in and out) coefficient Si to compute the Henry’s Law constant in the actual solution:

H

i =

Hio

Si

There is extensive literature on the Setschenow approach.4 The original work of Setschenow provides a method for calculating phase equilibria for systems for which conventional activity coeffi-cient models are unsuited. The

trick, however, is in determining the Setschenow coefficient for the gas of interest. This coefficient consists of two terms. There is a contribution from the molecular amine (a salting-in contribution given by kCa) and the various ions (salting-out contributions):

In(Si) = kC

a - 2.302585 h I

where I is the ionic strength. The system-dependent parameter h has anion, cation and gas contributions:

h = h_ + h+ + h

g

(Ionic strength is defined as I = ½∑m3z2

3, where m3 is the molal-ity of the salt ion and z3 is its charge number.)

The value of the salting parame-ter, k, depends on the specific hydrocarbon amine pair and is also temperature dependent. The salting-out part depends on the

1 2 3 4 5 6

Para

mete

r ‘b

’

Carbon number

1 2 3 4 5 6

Para

mete

r ‘a

’

Carbon number

Figure 1 How a and b parameters in the Setschenow (1892) salting-in coefficient correlate with carbon number for the n-alkanes. The value of “b” for n-butane is anomalous and is likely to be wrong

ogrt.indd 2 13/03/2013 11:21

C O M P R E S S O R S n T U R B I N E S n G L O B A L S E R V I C E

EBARA CORPORATION

www.elliott-turbo.com

n Customers: Global oil and gas producers.

n Challenge: Changing compression requirements as fields mature and production peaks.

n Result: Flexible, reliable compressor designs for extended, cost-efficient operations.

They turned to Elliottfor reliable compression solutions.

From Aberdeen to Rio de Janeiro, Calgary to Jakarta, Elliott compressors, turbines, and expanders deliver the reliable, efficient performance that producers and processors require. And every piece of Elliott equipment is backed by our unmatched global service network. Customers throughout the world turn to Elliott for critical turbomachinery and service because our resources are global and our execution is local. Who will you turn to?

The world turns to Elliott.

elliott.indd 1 12/03/2013 14:20

http://www.digitalrefining.com/134,sponsors,Elliott_Group.html#.UVAmUI6kJ94

campbell.indd 1 13/03/2013 09:55

http://www.eptq.com/jump.aspx?a=john&p=gas13

to interpolation, or even extrapola-tion, based on carbon number within a homologous series. This figure shows that, except for n-butane, which shows an abnor-mally low temperature dependence of the salting-in parameter (open data point), carbon number offers a reasonable basis for estimating parameters for missing components. Parameter a depends linearly on carbon number (salting-in becomes increasingly significant with increas-ing molecular weight of the solute gas), while b shows exponential dependence; that is, increasing temperature sensitivity.

A summary of the model’s corre-lation of the data presented in GPA RR-18013 and GPA RR-19510 for MDEA solutions is presented in Figure 2 as representative of the model’s quality. Figure 3 provides a similar comparison for H2 in MEA from Kling and Mauer’s 1991 data.8

It is not our intent to enumerate and discuss all the data for all the hydrocarbons and fixed gases in all the amines, but rather to present a flavour for how well a fairly gener-alised model based on the Setschenow (1892) salting-in and salting-out approach can be made to represent the experimental data; in this case, salting-in data. What we have found is that, by and large, this approach, combined with a good equation of state (EOS) model for the gas phase (for instance, Peng-Robinson or Redlich-Kwong-Suave), yields an excellent representation of the data. It also became apparent during model development that there are still significant voids in the data even for unloaded aqueous amine solvents. In particular, data for certain of the commercial amines are completely missing (DIPA specifically), and for most of the amines information on the tempera-ture dependence of solubility is scant where it exists at all (worthy subjects for research). In the next section, we shall see that when it comes to the effect of already dissolved acid gases — that is, acid gas loading — on hydrocarbon and fixed gas solubility, the data situa-tion is no better, and in some ways worse.


Salting-outJust as the amine tends to salt in non-polar compounds, a similar trend can be used for describing the salting-out effect that occurs from anions (h–) and cations (h+), with a possible contribution also from the soluble gas itself (hg). The model used in the ProTreat simula-tor was fitted to the form originally attributed to van Krevelen below, with pairing or mixing rules estab-lished that fit the scant data available:

log10

Hi = log

10 (1 ) = hI = (h

+ = h_ = h

g)I

Hio S

i

Model validationIn principle, accurate measure-ments of the low-pressure flash gas produced in a flash tank immedi-ately downstream from a high-pressure absorber should provide an excellent test of any model for hydrocarbon and inerts solubility in a treating solvent. All too often, however, such data are suspect because of the propensity of the solvent coming from the absorber sump to contain entrained gas. Any foaming tendency is likely to exacerbate such carry-under because foaming stabilises gas-

e6

e6

e7

e7

e8 e9

Henry

’s L

aw

Const

ant

– m

odel

Henry’s Law Constant – data

Toluene

Xylene

Benzene

Et-benzene

e8

e9

Figure 2 Goodness of fit of the Setschenow (1892) salting-in model to the GPA RR-180 solubility data for BTEX in MDEA from 25 to 120°C. Henry’s Law constant is in units of Pa/(mole fraction)

0.00010.0001

0.001

XH

YD

RO

GE

N –

mod

el

XHYDROGEN – data

25% mol% MEAWater

50% mol% MEA

0.01

0.001 0.01

Figure 3 Goodness of fit of the Setschenow (1892) salting-in model to Kling and Maurer’s (1991) solubility data for hydrogen in 25 and 50 mol% MEA

ogrt.indd 3 13/03/2013 11:21


other words, the solubility of the gases involved was roughly 1.45 times lower than would be predicted in the unloaded solvent — obviously, salting-out can be a huge correction and simply cannot be ignored.

The rich amine went to a zero duty flash drum operating at 106.5 psig. The flash gas make rate was estimated at 0.025 MMscfd at the plant site. Using the generalised model, ProTreat calculated 0.023 MMscfd. Table 3 compares the flash gas analysis with the calcula-tions of the generalised solubility model. For the most part, there is excellent agreement. It should be particularly noted that the CO2 loading of the rich amine from the flash drum was 0.69 at an amine strength of 21.4 wt%, so the ionic concentration was quite high and salting-out was severe. The VOC, HAP and fixed gas solubility model is very strong and extremely well qualified to estimate BTEX and most hydrocarbon and fixed gas solubilities with rather high accuracy.

measurements of the VOC and BTEX components present in the flash gas and degassed solvent in a carefully operated DEA plant. The data ranges in the following tables are shown as measured over an eight-hour data collection window. Table 1 shows inlet gas conditions.

Apart from the gas temperature (118°F, 48°C calculated by the ProTreat simulator versus 116–121°F, 46.7-49.4°C measured), nothing else concerning the treated gas was reported. However, the measured n-C6 and BTEX content of the rich amine leaving the absorber itself is compared with generalised model predictions in Table 2. With the exception of ethyl benzene, the generalised solubility model is in solid agree-ment with the data. The rich amine CO2 loading was 0.69 for this data set and, at this loading and bottoms temperature, the salt-ing-out parameter value is 1.45. In

liquid mixtures. It stabilises not just foams (structures of liquid films enveloping large volumes of gas), but also dilute emulsions contain-ing gas in the form of small bubbles dispersed in liquid. These emul-sions of entrained gas are carried from the column in the rich solvent and are separated in the down-stream flash tank. When this happens, it results in falsely high gas concentrations that are mistak-enly interpreted as dissolved gas. In fact, measured gas make rates and compositions almost always consist of dissolved plus some entrained gas, and such data can be quite erroneous and misleading. To add further fuel to the fire, many refinery amine systems contain multiple amine contactors that feed a common rich amine flash drum. When the flash gas is metered, it usually is not compensated for the effect of molecular weight and temperature.

Skinner, et al,12 made accurate

Temperature, °F 81–85Pressure, psig 1040–1110Gas flow, MMscfd 10.9–11.2Composition, mol%Water 0.00632 n-pentane 0.456CO

2 6.90 n-hexane 0.0700

Methane 83.30 Benzene 0.0344Ethane 6.41 Toluene 0.0213Propane 1.82 Et-benzene 0.00074i-butane 0.449 m-xylene 0.00404n-butane 0.530 o-xylene 0.00091

Inlet gas analysis

Table 1

Measured Generalised modeln-hexane nd–0.02 0.001Benzene 23–27 28Toluene 10–14 16.3Et-benzene 0.2–0.4 0.004m,p-xylenes 1.6–2.6 2.3o-xylene 0.6–0.9 0.51

n-C6 and BTEX in rich amine from

flash tank, ppmw

Table 2

Measured Generalised modelCO

2, mol% 32.5 33.98

Methane, mol% 60.6 59.72Ethane, mol% 4.4–4.8 4.82Propane, mol% 0.68–0.73 0.96i-butane, mol% 0.10–0.11 0.07n-butane, mol% 0.14–0.15 0.18n-pentane, mol% 0.03–0.04 0.073n-hexane, ppmv 162 66Benzene, ppmv 1000 900Toluene, ppmv 521 466Et-benzene, ppmv 17 12m,p-xylenes, ppmv 95 64o-xylene, ppmv 14 14

Flash tank make gas analysis

Table 3

Table 3

Inlet gas SolventTemperature, °C 33.3 Temperature, °C 43Pressure, barg 50.4 Pressure, barg 61.04Flow, kg/h 582 450 Flow, kg/h 900 000Components, mol% Components CO

2 5.83 CO

2 (loading) 0.024

C1 79.04 MDEA, wt% 35

C2 8.26 Piperazine, wt% 9

C3 1.58 Water Balance

iC4 0.13

nC4 0.11 Wash water

iC5 0.02 Temperature, °C 30

nC5 0.01 Pressure, barg 55.16

Benzene 0.91 Flow, kg/h 3044 N

2 3.98 CO

2, ppmw 1

H2O 0.13

Absorber inlet streams

Table 4

ogrt.indd 4 13/03/2013 11:21

Flexible H2S Removal

reliability. Over 200 licensees in 29 countries are meeting today’s compliance requirements by

turning nasty H2

and environmentally sound, LO CAT is the technology of choice for H2S removal / recovery.

LO CAT®

www.merichem.com

Sweet Solutions®

--

-

-

C

M

Y

CM

MY

CY

CMY

K

Merichem Company LO-CAT.pdf 1 3/4/2013 7:36:22 AM

merichem.indd 1 12/03/2013 14:38

http://www.digitalrefining.com/91,sponsors,Merichem_Company.html#.UVAmmI6kJ94


The tendency for benzene to accu-mulate in the inside of the column was studied by taking the converged column results from the initial “starved” solvent condition and perturbing the solvent flow in the ProTreat simulator from 900 000 kg/h to 1 000 000 kg/h using a Restart feature. (The Restart feature allows a previously converged column profile to be used as the starting conditions for a new simu-lation. By stopping the Restart simulation after various numbers of iterations along the path to conver-gence and capturing the results, an unsteady state path of the tempera-ture and concentration profiles could be studied.) While this case study does not necessarily represent the true time-dependent behaviour of the column, because there is not a direct quantitative correspondence between the number of column iterations and time, it does represent a path along which the column could be expected to behave with time.

Figure 4 summarises the progres-sion of the temperature profile over the arbitrary time intervals studied, while Figures 5 and 6 show the benzene concentration profile and the CO2 loading profiles, respec-tively, in the solvent inside the column over the same window. With the collapsing temperature profile, benzene is predicted to increasingly accumulate to quite high levels inside the column. This was somewhat counter to our expectations, which were derived from knowing that benzene salt-ing-in (ie, solubility in the amine) is greater at higher temperatures. Thus, higher temperatures mean benzene salts in more and, conversely, lower temperatures (which are what is seen here) should mean benzene salts in less (has lower solubility). In fact, at first glance, the opposite happens. It is not until the CO2 loading profile inside the column is exam-ined that this makes sense.

As Figure 6 shows, the benzene concentration explodes simultane-ously with the collapsing CO2 loading. No longer having a high CO2 loading in the middle of the column causes the salting-out effect

solvent loading at the rich end of an absorber.

A 4m-diameter LNG absorber with 16m of IMTP-50 random pack-ing was initially operating under the conditions shown in Table 4. The column was simulated to be producing higher-than-specified CO2 in the treated gas (1400 ppmv) caused by a slightly starved amine supply. Low amine flow resulted in a large, broad, internal temperature bulge.

Commercial implications ofsalting-outThe solubility of benzene in a piper-azine promoted MDEA solvent for CO2 removal in an LNG plant is used as an example of just how much salting-out matters. It is known that the solubility of benzene in MDEA is an increasing function of temperature.13 The question is the extent to which this higher solubility is mitigated by salting-out from the high salts content caused by high

20

10

15

25

0

30

5

40 50 60 70 80 90

Segm

ent

num

ber

from

top

Solvent temperature, ºC

35

Old steady state

New steady state

d1d2d3d4d5d6

Figure 4 Progression of temperature profiles from one steady state to another

20

10

15

25

0

30

5

0 0.02 0.060.04 0.08 0.10 0.12 0.14 0.16

Segm

ent

num

ber

from

top

Liquid benzene concentration, mol%

35

Old steady state

New steady state

d1d2d3d4d5d6

Figure 5 High benzene concentrations building up in the column

ogrt.indd 5 13/03/2013 11:21


9Kling G, Maurer G, Solubility of hydrogenin aqueous ethanolamine solutions attemperaturesbetween323and423K,J. Chem. Eng. Data,36,390-394,1991.10MokraouiS,ValtzA,CoqueletC,RichonD,Mutualsolubilityofhydrocarbonsandamines,GPA Research Report 195 (RR-195),2008.11Setschenow J, Ann. Chim. Phys., Über die Konstitution der Salzlosungen auf Grund Ihres Verhaltens zu Kohlensäure,25,226,1892.12SkinnerFD,ReifDL,WilsonAC,BTEXandotherVOCemissionsfromanaturalgasaminetreater,Topical Report to Gas Research Institute,GRI Project Number 5094-220-2796, Feb1996.13ValtzA, Guilbot P, Richon D,Amine BTEXsolubility, GPA Research Report 180 (RR-180),2002.

Nathan Hatcher joined Optimized GasTreating, Buda, Texas, as Vice-President,TechnologyDevelopment,in2009.HeholdsaBSinchemicalengineeringfromtheUniversityof Kansas and is currently a member of theAmineBestPracticesGroup.Email: [email protected] Jones joined Optimized Gas Treating,Inc as a Software Development Engineer in2012. HeholdsaBSinchemicalengineeringfrom McNeese State University and a MS inchemical engineering from the University ofNewMexico.Ralph Weiland is a co-founder of OptimizedGas Treating in Clarita, Oklahoma, US. HeholdsBASc,MAScandPhDdegreesinchemicalengineering from the University of Toronto,thenspenttwoyearsasapost-doctoralfellowin applied mathematics at the University ofWesternAustralia.Email: [email protected]

making predictions that can truly pass a rigorous test by regulatory authorities.

ProTreatisamarkofOptimizedGasTreating,Inc.DGAisamarkofHuntsmanCorporation.

References1BarrettPVL,Gasabsorptiononasieveplate,PhDthesis,CambridgeUniversity,1966.2Browning G J, Physical solubility of carbondioxide in aqueous alkanolamines via nitrousoxideanalogy,B.E.dissertation,Departmentof

ChemicalEngineering,UniversityofNewcastle,Newcastle,NSW,Australia,1993.3Carroll J J, Mather A E, A model for thesolubility of light hydrocarbons in water andaqueous solutions of alkanolamines, Chem. Eng. Sci.,52,545-552,1997.4HarveyAH,PrausnitzJM,Thermodynamicsof high-pressure aqueous systems containinggasesandsalts,AIChE Journal,35,4,635–644,1989.5Joosten G E H, Danckwerts P V, Solubilityand diffusivity of nitrous oxide in equimolarpotassium carbonate-potassium bicarbonatesolutions at 25°C and 1 atm, J. Chem. Eng. Data,17,452-454,1972.6JouF-Y;NgH-J,Critchfield JE,MatherAE,Solubilityofpropaneinaqueousalkanolaminesolutions, Fluid Phase Equilibria, [194–197],825–830,2002.7JouF-Y,MatherAE,Solubilityofpropyleneinaqueousalkanolaminesolutions,Fluid Phase Equilibria,[217]201–204,2004.8KlingG,MaurerG,TheSolubilityofhydrogeninwaterandin2-aminoethanolattemperaturesbetween323kand423kandpressuresupto16MPa, J. Chem. Thermodynamics, 23, 531-541,1991.

to shrink dramatically because the lower ionic salt concentrations asso-ciated with lower loading cause less desolubilisation. More importantly, these factors give benzene an increased incentive to enter the amine solution and to exhibit much higher levels than when CO2 is pres-ent; the temperature profile is collapsing towards the bottom of the tower. The behaviour of this column may go some way towards explaining the old saying, well known to those who have regularly operated amine units: “Amine units do not like rapid changes”. Any rapid change to feed gas CO2 or H2S content, amine temperature or amine flow rate that will rapidly swing the column’s internal temperature profile will also rapidly swing the amount of hydro-carbon stored inside the column. It is easy to imagine situations in which a sudden change in a varia-ble (for instance, an increase in the CO2 level in the gas or a higher gas rate) will lead to an equally sudden lowering of the equilibrium hydrocarbon content in a liquid that already contains hydrocarbon at a high pre-change level. In other words, the solubility limit will be exceeded locally. If the rate of hydrocarbon mass transfer from the liquid back into the gas occurs more slowly than liquid hydrocarbon forms, an organic liquid phase will coalesce, surface tension at the gas-liquid interface will drop, and the energy needed to expand the froth will fall dramat-ically. Thus, the froth on trays can be expected to turn to foam, which may eventually be ejected from the column.

ConclusionsA good hydrocarbon solubility model in amine systems will prop-erly take into account salting-out effects as well as salting-in effects. The VLE data for salting-out parameters (and, to a much lesser extent, salting-in effects) are scarce and the industry would stand to benefit significantly from the collec-tion of additional data, especially on the contribution of the sulphide and bisulphide forms of H2S. This will be especially important for

20

10

15

25

0

30

5

40 50 60 70 80 90

Segm

ent

num

ber

from

top

CO2 loading, m/m

35

Old steady state

New steady state

d1d2d3d4d5d6

Figure 6 ApproachoftheCO2loadingprofiletothenewsteadystate

ogrt.indd 6 13/03/2013 13:41

eliminate handoffs

With integrated natural gas processing technology from one supplier, UOP is the right connection.

Integration isn’t just the key to a successful contaminant-removal system.

It’s also fundamental to saving money as you monetize your resources. As

your single-source supplier for a wide range of gas processing technologies,

UOP reduces engineering and execution schedules so that your time to first

gas is reduced. UOP technologies are designed for true integration, delivering

energy savings. For removal of acid gas, water and trace sulfur compounds

plus recovery of natural gas liquids, don’t risk handoffs. Go with the

integrated processes of UOP.

For more information, visit www.uop.com/gasprocessing.© 2012 Honeywell International, Inc. All rights reserved.

C

M

Y

CM

MY

CY

CMY

K

UOP Gas Processing Eliminate Handoffs bleed.pdf 3/1/13 12:37:33 PM

uop.indd 1 12/03/2013 15:00

http://www.digitalrefining.com/149,sponsors,UOP___A_Honeywell_Company.html#.UVAmso6kJ94

Industrial gases support the natural gas production chain

For decades, industrial gas companies have assisted producers in the petrochemical

landscape to ensure they are able to deliver the levels of product quality and operational safety demanded by changing legislation, environ-mental regulations and customer requirements.

Industrial gases come into play from the very earliest upstream stage, where offshore exploration equipment is manufactured from steel and special metal alloys. These materials need to be cut, heated, welded and coated with the aid of industrial gases. In particular, many applications associated with drilling equipment now demand the use of higher performance materials. These materials are a world away from what was being used as recently as 10 years ago and, as modern material specifica-tions evolve, metal fabrication companies need to adapt their gases to suit the latest welding processes.

Offshore exploration equipment is often situated in some of the roughest and most inhospitable seas, and has to be able to with-stand enormous loads from huge waves and swells and highly corro-sive conditions. High-quality steel and the most up-to-date production methods are necessary for explora-tion operations under such harsh conditions. In this context, welding becomes particularly important, as the huge steel towers and support stilts are manufactured from many individual steel segments. A faulty weld seam on a single component can have catastrophic conse-quences. Cracks or dangerous

The contribution of industrial gases across the petrochemical processing spectrum is crucial to performance and safety

STEPHEN HARRISON and ERNST MIKLOSLinde Gases

salt-water corrosion could lead to a rupture in one or more of the steel components. Performance stand-ards for exploration equipment components, particularly those operating offshore, therefore confront manufacturers with tough challenges.

As more high-strength steels are being specified to manufacture increasingly tough drilling equip-ment, this has resulted in an extra preheating stage prior to welding in order to safeguard the metallur-gical properties of the steel. Preheating prevents failures, such as hydrogen-induced cracking, as well as common failures in the heat-affected zone. As more fine-grained structural steels are being used to construct apparatus and equipment for the oil and gas industry, the importance of preheating prior to the welding process is becoming a focal point.

Welding in this application is a complicated affair. To begin with, the thick metal pieces need to be

preheated. If this is not done, the large, cold steel plates will lose heat too quickly and the metal will not be completely melted in the weld-ing zone, making a secure connection impossible. Preheating will also prevent the build-up of cold cracks, which can occur due to hydrogen exposure or internal stress in the component.

This is particularly important when treating high-strength steels. After the weld, these materials must be post-heated for around two to three hours to diffuse any rogue hydrogen atoms in the weld seam. For manufacturers that have to maintain a fast production speed, it is vital that they quickly reach a preheated temperature of greater than 100°C.

In response, special burners for preheating steel before welding are required. Linde has engineered a Lindoflamm acetylene burners for preheating steel before welding takes place. Acetylene provides high heat intensity in the primary


2400

2300

Acetylene

2200

2100

2000

1900

1:4 1:8 1:12 1:16 1:20 1:24 1:28

Fla

me t

em

pera

ture

, ºC

Fuel gas / air ratio, m3/m3

1800

Methane

Ethene

Mixture with ethene

Acetylene

PropenePropane

Mixture with methyl acetylene

Figure 1 Flame temperatures of hydrocarbon gas/air mixtures

linde.indd 1 13/03/2013 11:42


helium is introduced into a “tri-mix” with oxygen and nitro-gen, or as heliox, which is helium mixed with low concentrations of oxygen.

These mixtures are critical to the well-being of the diver, as at these depths and pressures human blood responds to gas in a different way. The mixtures must be certified safe and need to avoid the presence of dangerous chemical molecules. The gas manufacturing process must be absolutely precise and quality controlled to ensure that no oils, methane or carbon monoxide infil-trate the mixtures. The effects of these substances are amplified at depth and their inhalation could prove fatal.

The human aspect of gas explora-tion is coming to the fore as gas companies are forced to go further afield to explore for natural gas deposits, frequently in places where the gas is difficult to extract. This means deeper exploratory dives and oil rig maintenance at greater depths.

Extraction and processingNitrogen (N2) and carbon dioxide (CO2) as bulk gases are used in huge quantities for the extraction and exploration of natural gas with induced hydraulic fracturing, commonly known as fracking. The traditional process for propagating hydrocarbons trapped in under-ground fractures had required high volumes of water, mixed with foaming agents and friction reduc-ers and injected at high pressure into the fractures, cracking open the shale and creating fissures, allowing gas or oil contained within them to flow freely.

However, some of the issues that have arisen over fracking have been linked to chemicals added to the water to assist in the fracking process. These chemicals are believed to have occasionally contaminated groundwater supplies, with documented cases of seepage into drinking water wells, often through improperly sealed or aban-doned drilling wells. There are also places where groundwater is only several hundred feet above gas reserves, where it is at risk of being

flame, establishing a focused flame, so that preheating occurs only in the weld area. This results in an increase in the speed at which the weld area is heated — as much as two-thirds faster than that achieved by other fuel gases — plus signifi-cant savings on total process cost.

As opposed to a propane gas flame, acetylene gas burns with a very precise, pointed “primary flame cone”, which drives the heat directly into the metal. Additionally, flame temperatures that can be reached with the associ-ated acetylene-compressed air torch — approximately 2400°C — are significantly higher than those achievable using other fuel gases in combination with air (see Figures 1 and 2).

Prolonging the lifespan of these offshore assets is another important factor, since equipment lifespan impacts the price of the final prod-uct. This issue has led to the development of special coating technologies that lengthen the life-time of offshore installations. Linde Gases Division has recently devel-oped a state-of-the-art cathodic protection technology as a first line of defence against metal corrosion.

Traditional arc spraying processes involve the use of air to coat metals. However, the enhanced Linspray arc spraying process employs a mixture of nitrogen and hydrogen to avoid oxidation of the applied coating. The new technique provides “active” protection, mean-ing the coating materials will actively repair the surface as it

detects corrosion. Results show it can improve the lifetime of heavy exploration and processing equip-ment by up to 50%. This is a significant step forward in reducing the number of maintenance inter-vals required and the associated costs involved.

A critical application of speciality gases within the gas extraction and processing industry is the testing of gas leakage detectors on offshore drilling platforms. With tonnes of natural gas being handled via the platform each day, any leakage could build up rapidly into an explosive atmosphere. Offshore platforms, therefore, have perma-nent monitors in operation, sniffing for gas seepage, and these gas detectors need continual testing and calibration with speciality gas mixtures.

Human elementDuring the natural gas exploration phase, a lot of the work takes place under the surface of the ocean, as specialised technical divers under-take oil and gas pipeline construction and maintenance, or maintenance of oil rigs and valves, and so on. Underwater, these divers breathe a range of different gases, depending on the depth at which they need to work.

Recreational divers breathe air, but commercial divers plunging to depths of up to 50 m need special-ised mixtures of oxygen and nitrogen in varying concentrations. When divers are required to go down to between 50 and 200 m,

1.6

1.2Acetylene

0.8

0.4

1:4 1:8 1:12 1:16 1:20 1:24 1:28

Fla

me p

ropagati

on r

ate

, m

/s

Fuel gas / air ratio, m3/m3

0

Methane

Ethene

Mixture with ethene

Propene

Propane

Mixture with methyl acetylene

Figure 2 Flame propagation rates for hydrocarbon gas/air mixtures

linde.indd 2 13/03/2013 11:43

more easily impacted by fracking. Conventional water treatment is often unable to remove the high concentrations of salts and other toxic and biologically disruptive compounds that can potentially be generated within wastewater from fracking — and if groundwater were to become contaminated, it could take years to clean an aquifer system.

Alternative techniques that can help mitigate water-related issues in fracking include employing CO2 mixed with alcohol or liquid nitro-gen (LIN) in a process known as dry fracking. The CO2/alcohol mix is also injected at high pressure underground to open up fractures, with the CO2 expanding as it vaporises, allowing natural gas to flow out through the cracks to be collected and processed.

Liquid nitrogen is used for dry fracking, a fracking process that has eliminated many of the problems associated with hydrofracking and could prove to be more acceptable to people concerned about the envi-


ronment. It uses very few, or no, chemicals and after fracking the nitrogen is released into the atmos-phere, which already comprises 78% nitrogen. Although relatively expensive in comparison to conven-tional chemical fracking, it is being used extensively for natural gas extraction in areas of high environ-mental sensitivity.

Air or pure oxygen as an indus-trial gas is used for sweetening or removal of sulphur compounds from LNG and LPG process streams. This is often done by employing a process called Merox from UOP. The Merox process uses catalysts and caustic soda to extract low molecular weight mercaptans from the refinery gas stream. The mercaptide-rich solvent must then be injected with oxygen from compressed air or industrial pure oxygen — for an enhanced regener-ation process — so that the mercaptides present are oxidised to disulphides. The disulphides are subsequently separated from the solvent so that typical mercaptan

levels in the gas product are controlled to less than 10 parts per million.

For refiners, throughput of natu-ral gas production can be limited by the speed at which plants can desulphurise natural gas. However, the more stringent the desulphuri-sation process becomes, increasing Claus plant loadings with hydro-gen sulphide and ammonia, the more frequent bottlenecks in the production process also become. Claus plants operating in refineries process concentrated hydrogen sulphide fractions, converting them into elemental sulphur. The tech-nology is also able to destroy pollutants, particularly ammonia.

Although not new, oxygen enrichment technology has now come to the fore as a viable and cost-effective solution for signifi-cantly increasing a plant’s sulphur handling capacity, as well as addressing problems associated with contaminants such as ammo-nia and hydrocarbons. Therefore, the use of oxygen debottlenecks the

Emission impossible

BCD Packing Ring -leading operators and OEMs confirm:

� Are you satisfied with the

effectiveness of your compressor

packings?

� Are you happy with leaking gas?

� Are you satisfied with the

durability of your packing rings?

� Are you comfortable with the

impact your compressor may be

having on the environment?

The BCD packing ring has proven its outstanding

performance in numerous applications worldwide.

Many compressor operators have installed the

HOERBIGER BCD packing ring and leading

OEMs have confirmed its high lifetime

and superior sealing efficiency. Their

compressors now run more efficiently

and with less fugitive emissions.

For extended lifetime, reduced

emissions and cost savings,

ask for the BCD packing ring:

[email protected]

www.hoerbiger.com

linde.indd 3 13/03/2013 11:43

http://www.digitalrefining.com/151,sponsors,Hoerbiger_Compression_Technology.html#.UVAm0I6kJ94


embrittlement (LME), which weak-ens the original structure of steel, aluminium and other metals in process plants, is the main threat. LME is a form of cracking that occurs when certain molten metals such as mercury come into contact with structural alloys. The most commonly affected materials include carbon steel, low-alloy steels, high-strength steels, 300 Series stainless steel and various alloys — nickel-based, copper, aluminium and titanium.

LME introduces an acute risk to an affected industrial plant. In a worst-case scenario, structural fail-ure could potentially result in a significant explosion in an oil refin-ery or LNG facility, resulting in catastrophic loss of life, excessive damage to capital equipment and long-term plant downtime.

When natural gas is shipped into a port, samples are taken not only determine the BTU value, but also the mercury levels, which, again, have a bearing on the function of turbines used in the transport process. To support this measure-ment, Linde produces a mercury gas standard typically between 1 and 60 µg/m3, which is the range at which mercury exists in natural gas.

Emissions control, monitoring and analysisCalibration gas standards are also widely used to monitor emission levels from gas processing or natu-ral gas-fired power plants. With the growing number of fracking opera-tions being initiated, natural gas is becoming a more cost-effective commodity, and an increasing number of natural gas power plants are coming online. Legislation requires close monitoring of emis-sions from these plants, focusing on the byproducts of the combustion process, sulphur dioxide, carbon monoxide and nitric oxide. Where coal-fired power station emission concentrations are in the 50 to 100 ppm range, natural gas has concen-trations around 10 ppm. However, this model is also associated with emissions such as ammonia and nitrous oxide, both of which are considered greenhouse gases, and

Accurate measurementIndustrial gases also play a major role, helping to determine the British thermal unit (BTU) content of natural gas, a measure of how much energy will be liberated when the gas is burned. Speciality gases’ calibration standards are made with known BTU values to measure LNG in the system, as this dictates how much the customer will pay for a specific volume of gas.

BTU values are very important when it comes to the transfer of ownership of a natural gas stream from one party to another. Natural gas is a very mobile commodity and can be moved over long distances, often across international borders. At the point where it changes ownership or crosses borders, the calorific value must be measured in an extremely accurate way. These massive quantities of gas are worth billions of dollars and the billing must truthfully reflect the energy value, consider-ing the high cost of error if the calorific value is not precisely measured.

Taxation authorities also base their revenue on the value of the product being distributed. The governments of countries involved in this trade have specified that the BTU measurement must be carried out using high-quality, high- accuracy, accredited gases, which are traceable to international meas-urement standards.

It is also essential to measure the amount of mercury in natural gas. In oil and gas production and processing plants, the mercury present in natural gas poses a formidable threat to the safety of humans and capital equipment, because of its propensity to amal-gamate with the materials of construction used for compressors and high-speed rotating equipment.

For instance, if mercury should amalgamate to pump system components or to the fins of turbine blades, it could cause considerable disruption by throwing these systems out of kilter, and therefore potentially cause immense struc-tural damage. Liquid metal

process and enables higher throughput and better returns on capital assets.

Oxygen enrichment of the combustion air significantly increases sulphur handling capac-ity. Associated benefits include increased productivity achieved without changing the pressure drop, more effective treatment of ammonia-containing feeds and less effort required for tail gas purifica-tion (reduced nitrogen flow). Oxygen enrichment is also a highly customisable approach to improv-ing Claus plant yield, with options varying from low-level oxygen enrichment to employing advanced proprietary technology to bring about capacity increases of up to around 150%.

TransportOnce the natural gas molecule has been located in the exploration phase, it must be transported on shore, either through natural gas liquefying stations or through pipelines.

Depending on the corrosive nature of the substance being trans-ported, pipelines are generally manufactured from high-value fabrication materials such as high-alloy or nickel-cladded steels. These are high investment materi-als, and fabricators are sensitive to the need to work with them appro-priately, and sophisticated industrial gases offer protection during the fabrication process. The smaller diameters and prefabricated steel pipes are joined using laser welding, and the bigger items are plasma or arc welded, all of them protected from corrosion through the use of purging gas to prevent deterioration. Gas purging is also environmentally friendly, allowing for a reduction in the use of chemi-cals to clean up an otherwise oxidised weld.

Where natural gas needs to be transported over long distances, usually by ship, liquefaction takes place, using refrigerant gases such as ethylene, before the lique-fied natural gas (LNG) is pumped into isocontainers or specially designed storage vessels on the ship.

linde.indd 4 13/03/2013 11:43


career in industrial gases spanning 20 years, he holds a master’s in chemical engineering from Imperial College, London.Ernst Miklos is Head of Segments Management, Manufacturing Industries, Linde Gases Division. He has 22 years’ experience in welding technology R&D and market development, and heads a global team of engineers focused on gas technologies to serve energy infrastructure. He has a degree in mechanical engineering from the Polytechnic Institute Timisoara, Romania, and is a certifi ed European Welding Engineer and a Member of the Research Board of the German Welding Society (DVS) and TWI, among others.

gases. Industrial gases will there-fore play a progressively critical role across the entire value chain, helping to ensure that all phases of the process operate effectively and safely, and that environmental impact is minimised.

Stephen Harrison is Global Head of Specialty Gases & Specialty Equipment with Linde Gases. He has worked in an international capacity for both Linde Gases and previously BOC, and now leads Linde’s Global Specialty Gases & Specialty Equipment business from Munich, Germany. A British Chartered Engineer (MIChemE) with a

measurement of these compounds is enabled with high-precision speciality gas mixtures.

When it comes to natural gas combustion control in power gener-ation, plant operators need accurate information about the combustion gases in the system. To avoid over-feeding or underfeeding natural gas into the system, operators rely on BTU values to provide insight and it is therefore imperative to have highly accurate calibration gases made with those specifi ed BTU values in order to monitor the combustion gases and help main-tain control throughout the process.

Inside the combustion chamber, it is also necessary to monitor oxygen levels to support fuel and carbon dioxide levels, to control the combustion process and minimise greenhouse gas emissions. These measurements also require special-ity gases calibration mixtures to support the instrumentation. Effective monitoring of combustion and optimisation of the process is the key element of emissions reduction.

High-purity “zero” gases are used to zero out the analytical instruments that monitor the output of fl ue gases from natural gas processing plants. Accurate and pure zero gases are a critical element of monitoring, because any trace amounts of impurities can alter or skew results.

An increasing number of mole-cules is being analysed and measured to support the growing trade in emissions from natural gas combustion. In the US, there is a strong drive to monitor all the greenhouse gases being emitted from a process, and this trend is likely to go worldwide, bringing under the spotlight emissions of gases such as nitrous oxide, ammonia, hydrogen chloride, formaldehyde, sulphur hexachlo-ride and carbon dioxide.

The futureAs the oil and gas industry advances, the more complex the processing of natural gas will become. This maturing value chain is going to demand increasingly high-quality and sophisticated

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

we timely deliver since over 35equipment packages tailored to meet specific customer’s requirements; for all kind of Oil

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 CO

linde.indd 5 13/03/2013 13:42

http://www.eptq.com/jump.aspx?a=prosernat&p=gas13

Selecting the best amine/solvent for gas treating is not a trivial task. There are a number of amines available to remove con-taminants such as CO2, H2S and organic sulfur compounds from sour gas streams. The most commonly used amines are monoethanolamine (MEA), diethanolamine (DEA), and methyldiethanolamine (MDEA). Other amines include diglycolamine® (DGA), diisopropanolamine (DIPA), and triethanolamine (TEA). Mixtures of amines can also be used to customize or optimize the acid gas recovery. Temperature, pressure, sour gas composition, and purity requirements for the treated gas must all be considered when choosing the most appropriate amine for a given application.

Primary Amines

Secondary Amines

Tertiary Amines

Mixed Solvents

Choosing the Best Alternative

Process Selecting the Best Solvent for Gas TreatingInsight:

bre.indd 1 12/9/12 13:50:23

http://www.digitalrefining.com/61,sponsors,Bryan_Research___Engineering.html#.UVAnJI6kJ94

Formulated solvent reduces shale gas processing costs

With the development of new drilling technology and hydraulic fracturing

techniques, global shale gas produc-tion has increased significantly over the past five years and is projected to continue increasing for several years. According to the International Energy Agency,1 “by 2020 the United States becomes a net exporter of natural gas and is almost self-sufficient in energy, in net terms, by 2035.” “…as demand increases by 50% to 5 trillion cubic meters in 2035. Nearly half of the increase in production to 2035 is from unconventional gas, with most of this coming from the United States, Australia, and China…” Increases in shale gas production and decreasing energy prices glob-ally will require competitive exploration and operational expenses among the energy options.

In early 2011, after having plant operational issues, poor reliability of gas treating facility equipment and high operational costs, a US gas treating operator met with Dow Oil, Gas & Mining to discuss options to reduce operational costs. Based on the operational perfor-mance information provided for newly released Ucarsol Shale Specialty Solvent from Dow, refer-ence facilities, solvent technical performance simulations and no requirement for expensive and complicated chemical additives, the operator decided to offer one unit that was about to start up as a demonstration facility. It was one of two similar units supplied with sour gas from the same gas pipe-line and located at the same facility.

A gas treating facility is upgraded to a formulated solvent to achieve reduced operating costs compared with operations using a commodity amine blend

ROBERT L DOTSON Dow Oil, Gas & Mining

Amine/CO2 chemistry

The absorption of CO2 from a gas stream uses the solvent’s ability to accept a proton, which provides a chemical driving force to assist solubilisation of the CO2. This propensity for the aqueous CO2 to act as an acid in the subsequent reaction with an amine takes place due to the higher dissociation constant (pKa) of the amine. The larger the difference in pKa between the amine and the CO2, the more complete the acid/base reaction that will take place. As Table 1 shows, MEA has the high-est pKa (9.5 at 25°C) and therefore

exhibits the highest degree of reac-tion of gas treating amines when compared to the pKa of aqueous CO2 of 6.0 at 25°C.2 In comparison, DEA has a pKa of 8.88 and MDEA has a pKa of 8.52 at 25°C.

In spite of its higher pKa compared to MDEA, DEA has disadvantages stemming from its propensity to degrade in the pres-ence of CO2 and corrode carbon steel.3 The degradation reactions of DEA are catalysed by, but do not consume, CO2. Work by Polderman and Steele,4 Kennard and Meisen,5,6,7,8 Blanc et al,9 and Kim and Sartori10 determined the DEA degradation reaction mechanism shown in Figure 1.

According to Kim and Sartori, the DEA-CO2 degradation reactions are initiated by the direct reaction of DEA and CO2 forming (HOCH2 CH2)2NCO2H (DEA-carbamate), which condenses to 3-(2-hydrox-yethyl) oxazolidone-2 (HEO). The HEO then reacts with another mole-cule of DEA, releasing a molecule of


pKa (25°C)

OH- 13.8Monoethanolamine(MEA) 9.5Diethanolamine(DEA) 8.88Methyldiethanolamine(MDEA) 8.52CO

2(firstdissociation) 6.0

Acid dissociation constant (pKa) of common amines and CO

2

Table 1

HO OH+ CO2

HN

DEA

HOHO

O

OH

OH

N

DEA-carbamate

HOHO

O

OH

OH

N

DEA-carbamate

HO

O

O

N+ H2O

HEO

HO

O

O

N + (HOCH2CH2)2NH

HEO DEA THEED

HOOHN

HO

NH

+ CO2

Figure 1 ReactionsofCO2catalyseddegradationofDEA

dow.indd 1 13/03/2013 11:46


a typical gas treating design:• 65 MMscfd feed gas treating capability with bypass• Absorber: 6ft diameter, 20 trays• Flash drum: 4.5ft diameter x 12ft TL-TL• Regenerator: 5ft diameter, 20 trays• Dual shell and tube L/R exchangers• Kettle reboiler/hot oil system from oil heater• 15 000 gallon system volume• 45 wt% Ucarsol solvent• Facility is manned during day operations• Facility is unmanned during night hours• Facility uses a DCS and paged alerts.

Trial descriptionUnit 1 had been in operation for four months using the operator’s standard 50 wt% MDEA/DEA blend prior to the completion of Unit 2’s construction. Prior to the addition of the Ucarsol solvent to Unit 2, the system was washed and drained to remove machining oils and particulates from construction, which can cause foaming and other issues. Unit 2 was then loaded with 45 wt% Ucarsol solvent and recom-missioned without issue.

Both Units 1 and 2 were loaded with sour gas feed to a point where they were able to operate satisfacto-rily and meet the treated gas specification. Since this operatorhas several units on the pipeline, it is able to move sour gas to these units, most of the time, to keep them loaded up to maximum feed rates. Unit 1, using the MDEA/DEA blend, continued to add the required chemical additives, as

CO2 and forming N,N,N’-tris (2-hydroxyethyl) ethyldiamine (THEED). Some of the THEED slowly condenses to form N,N’-bis (2-hydroxyethyl)piperazine (BHEP). Although several of the degradation products may contribute to the corrosivity of the solution, it is unknown which is the primary source of corrosion. Bicine, N-N-bis(2-hydroxyethyl)glycine, is another DEA degradation product that forms in the presence of oxygen or glyoxal (a dialdehyde used for H2S scavenging) in the amine.

Kim and Sartori demonstrated that these CO2-DEA degradation reactions are independent of the H2S concentration, are not initiated by thermal degradation, and are roughly proportional to the CO2 partial pressure. Blanc et al show that CO2 degradation products do not cause corrosion in gas environ-ments that contain a sufficientconcentration of H2S. It is generally agreed, per guidelines in API 945,11 that solutions with H2S-to-CO2 ratios of roughly 1:19 reduce corrosion rates substantially by forming a protective iron sulphide film.12 In a “CO2-only” natural gas stream, such as the Haynesville Shale region of the US, there is insufficient H2S concentration to form this stable layer of protective iron sulphide.

MDEA was developed to allow operations at higher amine concen-trations without CO2 degradation. However, unlike primary and secondary amines (MEA and DEA), MDEA has been shown to be kineti-cally limited for CO2 removal, as it does not form a carbamate molecule directly. MDEA is dependent on the CO2 hydrolysis reaction to create carbonic acid, which slowly dissoci-atestobicarbonateandfinallyreactswith MDEA for CO2 absorption. The

carbonic acid-to- bicarbonate reac-tion for MDEA is very slow compared to the direct reaction of the carbamate in the MEA and DEA reaction steps, resulting in a reduced CO2 reaction with MDEA,13 allowing CO2 to leave the absorber with limited removal (CO2 slip). Therefore, an activator is utilised as part of the MDEA formulation to effectively assist the reaction rate and tune CO2 selectivity, enhancing CO2 removal in the absorber. For a more detailed understanding of amine gas treating chemistry, refer to Fundamentals – Gas Sweetening, LRGCC.13

Solvent trialIn early 2011, the operator decided to compare Ucarsol solvent with its standard 80 wt% MDEA/20 wt% DEA blend diluted to a 50 wt% amine solution (40 wt% MDEA/10 wt% DEA/50 wt% water). The operator decided it would use the collected data to perform an economic evaluation and an opera-bility evaluation. Data collected during the trial are shown in Table 2.

Facility informationHaynesville shale gas composition for this area is shown in Table 3.

Unit 1 MDEA/DEA blendUnit 1 is a 250 gpm facility with a typical gas treating design:• 75 MMscfd feed gas treating with bypass• Absorber: 6ft diameter, 20 trays• Flash drum: 6ft diameter x 12ft TL-TL• Regenerator: 5ft diameter, 20 trays• APV plate and frame L/R exchanger• Direct-firedBryanreboiler• 10 500 gallon system volume• 50 wt% blended amine (80 wt% MDEA/20 wt% DEA)• Facility is manned during day operations• Facility is unmanned during night hours• Facility uses a digital control system (DCS) and paged alerts.

Unit 2 Ucarsol Shale H-102Unit 2 is a 200 gpm facility with

Table 2

Gas flow, MMscfd, daily avg.Filter changes, frequencyAmine make-up, gallonsChemical additives, total dollars (US)Antifoam use, pintsAmine analysis, bicine, wt%Amine analysis, iron, ppmv

Solvent trial data collection

Temp, °F 100 Pressure, psig 1000 CO

2, mol% 3.000

H2S, mol% 0.001

N2, mol% 0.100

Methane, mol% 96.400 Ethane, mol% 0.025 Propane, mol% 0.030 C

4+, mol% 0.444

100.000

Haynesville shale gas composition

Table 3

dow.indd 2 13/03/2013 11:47

ThyssenKrupp Uhde

ThyssenKrupp Uhde –Engineering with ideas.The key to our success is the creativity and resourcefulness of our employees. And it is this that keeps turning major challenges into solutions that are not only brilliant and innovative, but often set the standard for the entire engineering sector.

www.uhde.eu

http://www.digitalrefining.com/145,sponsors,ThyssenKrupp_Uhde.html#.UVAnQo6kJ94

recommended by the chemical supplier, to manage corrosion due to the DEA degradation products and bicine. Unit 2, using Ucarsol solvent, required no additives.

Amine samples for rich and lean amine streams were collected monthly and analysed in Dow’s amine lab, and the results evalu-ated. The evaluation used normal amine analysis to determine amine chemistry, amine concentration (wt%), bicine concentration (wt%) and iron (ppmw) for corrosion level potential, amine colour, amine clar-ity and amine contaminants.

Trial resultsUnit 1, using the MDEA/DEA blend, designed for 75 MMscfd of sour gas feed, was able to treat 70-80 MMscfd but was limited due to foaming issues at higher gas


rates. The bypass gas rate was typi-cally 5-10 MMscfd for a total sales gas rate of 75-90 MMscfd. Unit 2, using Ucarsol solvent, was designed for 65 MMscfd of sour gas feed, but due to the deeper CO2 removal and better operational performance at high gas rates the plant operators were able to load the facility with 90-100 MMscfd until equipment limits were met. A bypass gas rate of 15-25 MMscfd allowed a total sales gas rate of 105-125 MMscfd. The deeper CO2 removal allowed for a larger bypass gas rate. However, the H2S in the total gas stream was typically the limiting factor.

During the trial period, Unit 1 had three upsets due to severe foaming, causing the system to shut down automatically. A plant opera-tor was required to travel to the

facility, evaluate the cause and manually restart the system. Antifoam (8-16 ounces) was added until the foaming was under control. During these foaming events, some quantity of amine was lost through the treated gas outlet, and make-up MDEA/DEA blend was added to the system to adjust the concentration back to 50 wt%. Unit 1’s 20 micron rich and lean cartridge fi lters were changed twice during the trial. During this same period, Unit 2 had no major foam-ing issues, no shutdowns related to the amine and no fi lter changes. Antifoam was added to Unit 2, as needed, when minor foaming was thought to be occurring, but no serious foaming conditions occurred to shut down the unit during the trial. However, the unit did shut down due to boiler issues. Unit 2 used similar 20 micron cartridge fi lters. Both units ran approximately a 10% slipstream of the circulating amine through the particulate fi lters and carbon bed.

According to the plant operators, they “immediately noticed a differ-ence” in the operability of the two units. During most start-ups, Unit 1 was very onerous and often foamed. Operators felt like they were “fi ghting the unit” to get it started and keep it running. It required antifoam, corrosion inhibi-tor and the addition of bicine “killer”. Unit 2 started up with little issue, foamed occasionally with minor severity and required no chemical additives to manage corrosion.

The bicine concentration in 17 of the operator’s facilities using the MDEA/DEA blend and chemical additives for a 18-month period prior to the trial averaged ~8000 ppmw, with a maximum of 18 000 ppmw (see Figure 2). Over 18 months, the facilities using MDEA/DEA blend with chemical additives averaged about 3000 ppmw bicine. During the trial, Unit 1, operating with the MDEA/DEA blend with chemical additives, had bicine concentrations of 3200 and 4000 ppmw (two samples). After upgrad-ing to Ucarsol solvent, it has a non-detectable bicine concentration (<100 ppmv) after 16 months of

12000

20000

18000

16000

14000

10000

8000

6000

4000

2000

May

200

9

Dec 2

009

Jun

2010

Jan

2011

Jul 2

011

Feb 2

012

Aug 2

012

Bic

ine c

oncentr

ati

on,

ppm

w

0

Dec 2

009

Jun

2010

Feb 2

012

Aug 2

012

300

500

450

400

350

250

200

150

100

50

May

201

1

Aug 2

011

Nov 201

1

Mar

201

2

Jun

2012

Sep 2

012

Bic

ine c

oncentr

ati

on

, p

pm

w

0

Figure 2 Bicine concentration at 17 facilities with “CO2-only” gas using MDEA (40 wt%)/

DEA (10 wt%)/water (50 wt%)

Figure 3 Bicine concentration at 10 facilities using Ucarsol (45 wt%)

dow.indd 3 13/03/2013 11:47

www.amacs.com • 24hr EMERGENCY SERVICE • (281) 716-1179

• Mesh numerous alloys

& plastics

• Vane (Chevron)

• Multi Pocket Vane

• Insertion Mistfi x®

MIST ELIMINATORS

DISTRIBUTORS & SUPPORTSManufactured to customer specifi cations or engineered to meet performance requirements.

COALESCERS• Oil water separations • Haze removal from fuels • Removal of tower wet refl ux • Caustic treater applications

STRUCTURED PACKINGWoven, sheet metal, and knitted structured packing. Built to spec or performance requirement.

RANDOM PACKING

• Catalyst bed supports • Basket strainers • Nozzles • Outlet/Inlet baskets • Distributors • Hub and header laterals

WEDGE WIRE SCREENS

Wide variety of random packing types, sizes and materials in stock! Ask us how our patented SuperBlend™ 2-Pac can increase your capacity or effi ciency.

Valve (numerous options)Sieve or perforated

Bubble cap traysCartridge trays

Dual fl owBaffl e

From Engineering Through FabricationFor top performing mass transfer and tower internal components let

AMACS Process Tower Internals use your existing drawings or modify them to

improve your process. Our experienced in-house engineering and full scale

fabrication capabilities can streamline your project from inception

to installation.

Member of Fractionation Research

Visit our new website atwww.amacs.com

37_amacs.indd 1 12/03/2013 14:25

http://www.digitalrefining.com/56,sponsors,AMACS_Process_Tower_Internals.html#.UVAnXI6kJ94


has averaged 7.8 ppmw, with a peak of 48 ppmw, which quickly dropped back to the average concentration. Unit 1 had a pipeline brine-water incursion in July 2012, which brought chlorides (along with other contaminants) into the amine system. This resulted in the iron increasing to 11 and 14 ppmw until the 13 500 ppmw concentra-tion of chlorides was removed from the Ucarsol solvent using the Ucarsep Reclamation System from Dow.

Although this trial was scheduled for three months, the operator decided to remove the MDEA/DEA blend in Unit 1 after just over two months. The trial had provided enough information for the operator to conclude that Ucarsol solvent was their solvent of choice for the future. As stated by the operator, “Unit 1 did not perform as well as Unit 2”. Even though both units were treating the same gas, Unit 1 had more foaming events, shut-downs and filter changes. The chemical additive cost, higher bicine concentrations and therefore corro-sion potential, higher iron concentrations, and more solids filtered out of the amine (inter-preted as corrosion products by the operator) was enough to justify the solvent upgrade. Unit 2 ran cleaner (fewer solids filtered), which the operator deduced was a result of lower corrosion rates due to no bicine and degradation products in the solvent. Their cost evaluation also determined, based on their esti-mates for amine make-up usage and maintenance, that Ucarsol solvent was the lowest long-term cost of operation option rather than their standard MDEA/DEA blend with chemical additives. The lower price of the MDEA/DEA blend was offset by the cost of the required chemical additives and high mainte-nance costs. The operator’s cost analysis did not include the addi-tional cost of filter changes and lost revenue due to down time (as it was able to switch gas to other units, so it was not technically “lost revenue”), which would have increased the cost advantage. However, for a single or isolated unit that cannot switch gas to

may not have been completely drained and washed, and that some of the high bicine amine was in isolated piping (spare pump, bypass, and so on), contaminating the solvent when it was opened.

Iron concentrations for Unit 2 during the trial were measured at 3 ppmw and 3 ppmw, indicating very low corrosion rates (see Figure 4; Units 1 and 2 are indicated by the green and red lines). During the past 18 months, with the addition of eight more units (the total 10 shown as lines) upgraded to Ucarsol solvent, the iron concentration

operation. Unit 2, while operating on Ucarsol solvent, had a non- detectable bicine concentration during the trial and is still non-de-tectable after 18 months of operation (see Figure 3). All facili-ties operating on Ucarsol solvent have shown non-detectable (<100 ppmw) bicine (appearing as one line) with the exception of one facil-ity, which started near 100 ppmw and spiked up as high as 460 ppmw for a short period before dropping back down to non-detectable. No specific reason for this was found, but it was suspected that the system

New facility cost comparison 45 wt% Ucarsol 50 wt% MDEA/DEA blendAmine, initial fill, lb 58 000 64 500 Amine make-up, lb/yr 17 400 35 000 First year amine, lb 75 400 99 500 Chemical additives and antifoam, $ 720 24 912Equip maintenance due to corrosion, $ 0* 20 000Payback, yr 0.36 * No additional equipment maintenance expenses due to corrosion

Operational cost comparison of MDEA/DEA blend and Ucarsol: new 300 gpm facility

Table 4

Existing facility cost comparison 45 wt% Ucarsol 50 wt% MDEA/DEA blendAmine, Initial fill, lb 58 000 0 Amine make-up, lb/yr 17 400 35 000First year amine, lb 75 400 35 000Chemical additives and antifoam, $ 720 24 912Equip maintenance due to corrosion, $ 0 20 000Payback, yr 1.7

Operational cost comparison of MDEA/DEA blend and Ucarsol: existing 300 gpm facility

Table 5

30

50

45

40

35

25

20

15

10

5

Feb 2

011

Jun

2011

Sep 2

011

Dec 2

011

Apr 201

2

Jul 2

012

Oct 2

012

Iron c

oncentr

ati

on,

ppm

w

0

Trial period

Unit 2

Unit 1

Figure 4 Iron concentration for 10 facilities on Urcasol

dow.indd 4 13/03/2013 11:47


These cost estimates and payback calculations do not take into account the lost revenue for unit down time or additional opera-tional expenses, such as filter cartridges, which would further reduce the payback time. The reve-nue of any additional gas processed, total sweet gas sold to the pipeline (treated and bypassed), was also not included.

The operator also estimated first-year and annual operating expenses for an existing 300 gpm facility upgraded from MDEA/DEA blend to Ucarsol solvent, based on MDEA/DEA blend operational experience and trial experience (see Table 5). An existing facility would require the removal of the MDEA/DEA blend and an initial fill of the solvent. All other costs — make-up amine, antifoam, chemical additives and the estimated cost of corrosion maintenance — remain the same for the cost comparison. The cost of

“CO2-only” gas, the operator esti-mated equipment repairs to pumps, valves and piping due to corrosion (corrosion maintenance) at $20 000/year for the MDEA/DEA blend. The operator concluded that the payback is 0.36 years for a new 300 gpm facility using solvent in place of their standard MDEA/DEA blend.

Note: due to pricing confidential-ity of amine and antifoam, specific amine and antifoam costs are not provided. All operational cost comparison tables provide the pounds of amine for a system fill, estimated pounds of annual make-up amine, estimated annual cost of antifoam, operator-provided cost of annual chemical additives, and operator-estimated cost of corrosion maintenance in a MDEA/DEA system for a 300 gpm facility. Payback calculations are based on antifoam and amine costs at the time of the trial.

another unit, this lost revenue would further increase the cost advantage.

Operations cost comparisonThe operator estimated first-year and annual operating expenses for a new 300 gpm facility, based on MDEA/DEA blend operational experience and trial data (see Table 4). A new facility requires an initial fill of either amine. The annual operating cost of the MDEA/DEA blend requires the addition of chemical additives at $22 752/year for a unit this size. MDEA/DEA make-up amine, based on operator experience and trial results, was estimated to be two times the make-up requirement for Ucarsol solvent. Antifoam usage for the MDEA/DEA blend, based on the trial, was estimated at three times that for the Ucarsol solvent. From 10 years of operational experience with MDEA/DEA blends in

45 wt% Ucarsol 50 wt% MDEA/DEA blendGas flow, MMscfd • Unit 2, designed for 65 MMscfd treated up to • Unit 1, designed for 75 MMscfd, was able to treat 75-80 MMscfd 90-100 MMscfd during the trial and for 18 months during the trial. since. • Limited by severe foaming at higher sour gas rates. • Unit 2 treated gas to a lower CO

2 concentration,

allowing for increased sour gas bypass flow. Result: more sweet gas to sales at pipeline specifications.

Filter changes, • No filter changes during the trial. • Required two filter changes.total number

Amine • Required normal amine make-up volume during • Required amine make-up because of two severe foaming events.make-up, qty the trial.

Chemical additives, • Requires no additives. • Requires continuous addition of corrosion inhibitor and bicinetotal dollars (US) “killer” additives. • Chemical additive cost: $22 752/yr.

Antifoam use • Low antifoam usage due to a low foaming tendency. • Used three times on Unit 1 due to multiple foaming events and two severe foaming events.

Bicine concentration, • Unit 2 had non-detectable bicine concentration • Unit 1 bicine concentrations were 3200 and 4000 ppmw (twowt% (<100 ppmv) during the trial and after 18 months of samples). operation. • The bicine concentration for an 18-month period prior to the trial • Unit 1, after upgrading from MDEA/DEA blend to averaged 8000 ppmw, with a maximum of 18 000 ppmw. Ucarsol solvent, has a non-detectable bicine • Over the past 18 months, the MDEA/DEA blend has averaged concentration after 16 months of operation. ~3000 ppmw bicine. Iron concentration, • During the trial, Unit 2 maintained iron concentrations • Units operating on a MDEA/DEA blend from June 2009 to Novppmw between 1-4 ppmw, indicating very low corrosion rates. 2012 averaged 18 ppmw iron. Typical iron range is 8-20 ppmw.

• During the past 18 months, all 10 units upgraded to Ucarsol have averaged 8 ppmw, with a peak of 48 ppmw.

Operational reliability, • 100% amine operational reliability during the trial. • 95% amine operational reliability during the trial. Unit 1 shut% of time treating gas Unit 2 shut down due to boiler issues. down twice during the trial due to severe amine foaming events.

Trial data evaluation

Table 6

dow.indd 5 13/03/2013 11:47


References1 International EnergyAgency, World Energy Outlook 2012.2 Martell A E, Smith R M, Motekaitis R J,CriticallySelectedStabilityConstantsofMetalComplexesDatabase,NISTStandardReferenceDatabase 46, Version 7, US Department ofCommerce,2003.3 Kohl A L, Nielsen R B, Gas Purification,FifthEd,GulfPublishingCompany,1997,Ch3,SolutionDegradation,232-242.4 Polderman L D, Steele A B, Whydiethanolamine breaks down in gas treatingservice, Oil and Gas Journal, Vol 54, 30 July1956,206-214.5 Kennard M L, Meisen A, Control DEAdegradation,Hydro. Process.,Vol59,No.4,April1980,103-106.6 Kennard M L, Meisen A, Gaschromatographic technique for analyzingpartiallydegradeddiethanolaminesolutions,J. of Chromatography, Vol267,1983,373-380.7 Kennard M L, Meisen A, Mechanisms andkineticsofdiethanolaminecorrosion,Ind. Eng. Chem. Fundamentals,Vol24,1985,129-140.8 Kennard M L, Meisen A, DEA degradationmechanism, Hydro. Process., Vol 61, No 10,October1982,105-108.9 Blanc C, Grall M, Demarais G, The Part Played by Degradation Products in the Corrosion of Gas Sweetening Plants Using DEA and MDEA,Laurence Reid Gas Conditioning ConferenceProceedings,UniversityofOklahoma,NormanOK,1982.10KimCJ,SartoriG,Kineticsandmechanismof diethanolamine degradation in aqueoussolutions containing carbon dioxide, Int. J. of Chem. Eng. Prog.,October1984,64-71.11American Petroleum Institute (API),Avoiding Environmental Cracking in Amine Units, APIRecommendedPractice945,1stEd,August 1990, American Petroleum Institute,Washington,DC.12Nielsen R B, Lewis K R, McCullough J G,Hansen D A, Controlling Corrosion in Amine Treating Plants,LaurenceReidGasConditioningConference Proceedings, University ofOklahoma,Norman,OK,1995.13Seagraves J Quinian M, Conley J,Fundamentals — Gas Sweetening, LaurenceReidGasConditioningConferenceProceedings,UniversityofOklahoma,Norman,OK,2010.

Robert L (Rob) Dotson is a Senior GasConditioning Specialist with Dow Oil, Gas& Mining, with over 32 years’ experiencedesigning, operating and managing world-scalefacilitiesinthechemicalandoilandgasindustries.Heprovidestechnicalconsultationforrefineries,naturalgasandunconventionalgas treating facilities, natural gas liquids(NGL)treatingfacilitiesandpowerplantCO

2

capture(CCS)facilitiesaroundtheworld.Heholdsadegree inchemicalengineeringfromthe University of Florida, and is a licensedProfessionalEngineerinTexas.

make-up amine, reducing operating costs.

The higher frequency of filter changes in the MDEA/DEA system is likely due to the degradation products and bicine causing a weakening of the iron carbonate protective layer. Since upgrading the amine in the two units, filter changes are now once every two to three months rather than monthly or shorter. Operating costs due to these changes have been reduced too (the difference in filter change expense was not used in the opera-tional cost comparisons). Units 1 and 2’s facility corrosion mainte-nance has also been reduced over the past 18 months since the upgrade according to the operator, further reducing operational costs.

The operator has been able to increase treated gas production through the two units due to the lower sweet gas CO2 concentration produced after the upgrade. More sour gas can be processed in the gas treating system to a lower CO2 concentration, allowing more sour gas to be bypassed and blended to meet pipeline gas specifications. This increase in revenue was not included in the operational cost comparisons, but would further reduce the operational costs and payback time. However, the results for another facility can vary widely, depending on the facility and equipment condition, sour gas, and many other conditions and factors.

The operator determined the payback for a new 300 gpm facility is 0.36 years and the payback for an existing 300 gpm facility is 1.7 years. This cost evaluation is supported even when the estimates of higher antifoam and make-up rates are removed and maintenance corrosion costs eliminated, result-ing in a payback of 0.55 years for a new 300 gpm facility.

Evaluating the data collected during the two-month side-by-side trial shows the advantages of Ucarsol solvent over a MDEA/DEA blend (see Table 6).

UCARSOLandUCARSEParetrademarksofTheDowChemicalCompany.

the removed amine disposal was not included, as the operator was able to move the amine to another facility using the MDEA/DEA blend. The first-year expenses, in this case, are higher for the solvent to pay for the amine conversion, but the expenditure is recovered by the higher annual cost for the MDEA/DEA blend for antifoam, chemical additives and the esti-mated cost of corrosion maintenance. The operator concluded that the payback for the existing case is 1.7 years (see Table 5).

ConclusionsBased on the experience of the operator, who has used a 80 wt% MDEA/20 wt% DEA blend for over 10 years, the data collected during the trial and over 18 months of operation, it was concluded that the long-term operational cost of Ucarsol solvent has a payback that supports its use in their facilities. Through the end of 2012, the opera-tor has upgraded 10 facilities and continues to find opportunities to upgrade additional facilities to Ucarsol solvent to reduce operating costs.

The requirement and cost for the continuous addition of chemical additives (corrosion inhibitors and bicine “killer”) to control corrosion from DEA degradation products and bicine versus no chemical addi-tives for the Ucarsol solvent was a significant deciding factor for the operator. The trial results and over 18 months of operation indicating non-detectable bicine and low iron concentrations in the Ucarsol solvent further justified the upgrades for the operator.

Foaming in Unit 1 with the MDEA/DEA blend was an issue, especially when the system was loaded heavily with sour gas feed. This could have been seen as just a result of the system/equipment design, since the two units in the trial were not identical; however, Unit 1 has performed better since upgrading to Ucarsol solvent for 18 months and is now treating up to 110 MMscfd without significant foaming issues. Less severe foam-ing events have resulted in less

dow.indd 6 13/03/2013 11:47

Weir Minerals Lewis Pumps is the world standard for manufacturing pumps and valves in the sulphur chemicals industry. Offering a family of Lewis® steam-jacketed sulphur pumps, outstanding reliability in high-temperature sulphuric acid applications, and a complete family of valves, Weir Minerals Lewis Pumps continues a long tradition of offering superior products and services.

With its strong commitment to service, Weir Minerals Lewis Pumps can ship standard replacement wear parts in 72 hours to most international airports in emergency situations. With a complete line of pumps and valves, Lewis® Pumps products are right on target. Customers in more than 100 countries can’t be wrong.

Lewis® Pumps products are right

on target

ExcellentMineralsSolutions

LEWIS® PUMPSVertical Chemical Pumps

8625 Grant Rd.St. Louis, MO 63123T: +1 314 843-4437F: +1 314 843-7964

Email: [email protected]

Copyright © 2011, EnviroTech Pumpsystems, Inc. All rights reserved. LEWIS and LEWIS PUMPS are registered trademarks of Envirotech Pumpsystems, Inc.; WEIR is a registered trademark of Weir Engineering Services Ltd.

lewis.indd 1 12/03/2013 14:41

http://www.eptq.com/jump.aspx?a=Weir&p=gas13

z4412.indd 1 3/11/2013 9:26:40 AMlinde.indd 1 12/03/2013 16:38

http://www.digitalrefining.com/139,sponsors,Linde_Engineering.html#.UVAnmI6kJ94

Qatar’s LNG industry

Natural gas is the most impor-tant energy source for the future. It supplies around

one-fifth of the world’s energy needs, compared with one-third from oil and one-quarter from coal, and will continue to play an impor-tant role in meeting demand for energy because of its abundance coupled with its environmental soundness and multiple applica-tions across all sectors.

Liquefied natural gas (LNG) is expected to play an increasingly important role in the natural gas industry and global energy markets in the next several years. The combination of higher natural gas prices, lower LNG costs, rising gas import demand and the desire of gas producers to monetise their gas reserves is setting the stage for increased global LNG trade.

LNG is natural gas that is stored and transported in liquid form at atmospheric pressure at a tempera-ture of -260ºF (-160°C). In international trade, LNG is trans-ported in specially built tanks in double-hulled ships to a receiving terminal, where it is stored in heav-ily insulated tanks. The LNG is then sent to a regasifier, which turns the liquid back into a gas that enters the pipeline system for distribution to customers as part of their natural gas supply.

The low density of natural gas makes it more costly to contain and transport than either oil or coal. Due to its special processing and handling requirements, the costs of moving natural gas are significantly higher than the costs of moving oil. And the relative costs of moving gas or oil by pipeline or by tanker differ

LNG provides an economic means to transport natural gas over long distances. Resources and technology have enabled Qatar to become a leading exporter

MEGHNA BAHL and SANDEEP KUMAR Fluor

substantially too. This influences regional inter-fuel competition and thus natural gas markets. Although pipeline costs increase linearly with distance, LNG — requiring liquefac-tion and regasification regardless of the distance travelled — has a high threshold cost but a much lower increase in costs with distance. Thus, shorter distances tend to favour pipelining, whereas longer distances favour LNG.

The marketGrowth of trade in LNG is driven by increasing demand and declining domestic natural gas resources in gas-consuming countries, and by the desire of gas-producing countries to commercialise their resources.

Due to the low weight of interna-tional trade in total production, there is not a globalised natural gas market, but rather regional markets, which vary in terms of their organi-sation, maturity and market structures. There have been three distinct and relatively independent markets for LNG, each with its own pricing structure. Price risk is inher-ent in each pricing structure, although the degree of risk differs among the markets. LNG prices are benchmarked to competing fuels.

The countries of North America constitute a very integrated and mature market for natural gas. The North American natural gas market is almost self-sufficient. In the US, the competing fuel is pipeline natu-ral gas, and the benchmark price is either a specified market in long-term contracts or the Henry Hub15 price for short-term sales. Importers and exporters involved in US LNG transactions are exposed to a

significant level of risk given the high degree of price volatility in US natural gas markets.

In Europe, LNG prices are related to the prices of competing fuels such as low-sulphur residual fuel oil. However, LNG is now starting to be linked to natural gas spot and futures market prices.

In Asia, prices are linked to imported crude oil. The pricing formula typically includes a base price indexed to crude oil prices, a constant and perhaps a mechanism for the review/adjustment of the formula. Asian prices are generally higher than prices elsewhere in the world.

Market and price convergence will be possible not only if LNG trade grows sufficiently, but also if the short-term/spot market sees marked growth too. While it may seem logical that more LNG should begin to be traded under short-term conditions in the interest of economic efficiency, the noted financial capital risk of investors may compel them to maintain longer terms in an effort to reduce their risk exposure.

In 2011, global LNG trade grew by 20.7 million t/y, a growth of 9.4% compared with 2010. The main contribution to the increase of LNG flows came from Qatar, as the country was responsible for 67% of additional LNG produced in 2011. LNG consumption in Asia contin-ued to grow strongly (+14.8%), reaching a total of 153 tonnes, which is about 63.6% of the world’s LNG trade.

Until 2014, the LNG market will grow by 2% due to the limited number of liquefaction projects


fluor.indd 1 13/03/2013 12:02


reserves of natural gas exceed 25 trillion cubic metres, about 14% of the world’s total and the third larg-est. Qatar is the world’s largest exporter of LNG, supplying around one-quarter of global exports in 2010 (see Figure 1).

The nation’s economic policy has focused on developing its vast oil and natural gas reserves with considerable success. Its transfor-mation began in 1991, when the Government of Qatar and Qatar Petroleum committed to construct the largest LNG port in the World for ~$1 billion dollars at Ras Laffan. Incidentally, this investment deci-sion occurred prior to any firm buyer commitments to purchase LNG. The port, originally designed for ~50 MTA of LNG, will be doubled in capacity over the next few years to handle all of Qatar’s expansion plans, including ~100 MTA of LNG, as well as conden-sate, LPG and gas-to-liquids (GTL) products.

Qatar has the advantage of having resources concentrated in one field. The gas reserve is contained in the single largest discovered non-associated gas field, called the North Field, and is esti-mated to contain some 900 trillion cubic feet of reserves. For compari-son, this is over 60 times larger than Exxon Mobil’s Indonesian LNG operation, which was devel-oped over 30 years from a reserve base of 14 trillion cubic feet.

Perhaps the most significant advantage that Qatar has, beyond its natural resources, is its geographic location. Strategically positioned almost mid-way between the Pacific Basin and Atlantic Basin, it has the ability to send cargoes to ports ranging from the US Gulf Coast to Korea. This advantageous positioning will allow the country to pit consumers against one another and fetch globally higher pricing for its shipments. Of course, demand restraints vary between the basins, but the underlying needs are the same regardless of the way contracts are priced and what they are tied to. Competition will breed synergy and Qatar is in the ideal location to exploit the emerging market.

going into operation during this period.

The impact of the nuclear crisis in Japan has boosted its LNG imports to 8.5 million t/y during 2011. Also, Japan has imported large volumes of spot LNG at very high prices to ensure reliability of supply. During 2012, Japan was expected to increase its LNG spot requirements even more due to the nuclear shutdown.

China is emerging as a main growth centre in terms of invest-ment in regasification plants. Hence, it will be importing large volumes of LNG in the future. In the next two years, China will provide five new plants with a total capacity of 14 million t/y.

LNG supply is expected to expand after 2016 due to a variety of factors:• Ongoing projects in Australia, which are set to add about 33 million t/y of LNG to the current supply from 2016• Increasingly, LNG cargoes initially meant for the US have been diverted to Asia and Europe in line with discoveries of shale gas in the US, which have turned it into a major exporter, instead of a net importer, of gas• In parallel, other regions in the world, including China, Australia and Europe, could tap their uncon-ventional gas reserves

• Development of floating LNG (FLNG) production, where offshore gas fields could be accessed at a lower cost and with a smaller envi-ronmental footprint. Some industry players such as Shell have already invested in this development. The success of FLNG could bring even more LNG supply projects online, and open up more supplies and choices for LNG buyers.

Qatar – supplier of LNGQatar is richly endowed with natu-ral resources; the oil and gas sector dominates the economy. Despite the global financial crisis, Qatar has prospered in the last several years. In 2010, the country had the world’s highest growth rate. Qatari authorities throughout the crisis sought to protect the local banking sector with direct investments into domestic banks. Economic policy is focused on developing Qatar’s non-associated natural gas reserves and on increasing private and foreign investment in non-energy sectors, but oil and gas still account for more than 50% of GDP, roughly 85% of export earnings and 70% of government revenues. Oil and gas have likely made Qatar the highest per-capita income country with the lowest unemployment. Proven oil reserves of 25 billion barrels should enable continued output at current levels for 57 years. Qatar’s proven

15

25

20

10

5

Qatar

Oman

Yem

en

Indone

sia

Mala

ysia

Austra

lia

Russia

Brune

i

Nigeria

Algeria

Egypt

Equato

rial G

uinea

Norway

PeruUSA

BelgiumLib

ya

Trini

dad an

d Tobag

oUAE

Pro

po

rtio

n o

f to

tal 2

010,

%

0

Middle East

Asia Pacific Atlantic

Figure 1 Qatar leads in LNG exports Source: BP 2011

fluor.indd 2 13/03/2013 12:02

An important factor enabling Qatar LNG to compete worldwide, particularly in the liquid UK and US markets, is the advancement in large LNG ship technology. It has actively developed large LNG tank-ers in a range from 210 km3 (Q-Flex) to as large as 265 km3 (Q-Max). This will lead to signifi -cantly lower shipping costs, which will allow Qatar to expand its global reach and compete long term.

With each new expansion train, Qatar has set the record for the largest train ever built. The Qatar Gas train 7 was the last in its expansion plan to reach a capacity of 77 million t/y of LNG exports.

To help accomplish its vision, Qatar Petroleum has selected strong partners for developing the North Field gas resource. Qatar has strong relationships with Japan, Korea, the US and Europe. Multinational oil companies includ-ing Exxon Mobil and Total currently participate in LNG joint ventures, along with several


Japanese trading companies. In addition, new relationships have been developed with players such as ConocoPhillips and Shell to capture US LNG and GTL market opportunities.

Australia may emerge as a threat in terms of production of LNG. If all projects currently under construction or at a proposal stage are developed, Australia may produce up to 98.8 million t/y in 2016/17. Australian projects are facing major challenges that will determine the success or failure of each project as well as delaying entrance into operation. Expensive manpower together with stringent immigration laws have led to cost increases of up to 30% in some cases.

ConclusionAs the global economy continues to grow and evolve, the role of energy will become ever more crucial and, as its importance is magnifi ed, abil-ity to deliver energy will be vital. As traditional sources of gas supply

fail to meet rising demand, LNG will have an opportunity to experi-ence momentous growth.

LNG presents an opportunity for the monetisation of plentiful stranded natural gas resources. The ability to move gas from displaced markets to high-demand regions has revolutionised natural gas markets. It will be growth in inter-national LNG trade that continues to globalise gas markets.

References1 LNG Global Market Outlook, 2011.2 GIIGNL - LNG Industry Report, 2011.

Meghna Bahl is a Control System Lead with Fluor and has more than six years’ experience in the oil and gas industry. She holds a degree in instrumentation engineering from IIE, Kurukshetra University.Sandeep Kumar is a Project Engineering Manager with Fluor and has more than 13 years’ experience in process design for gas processing, refi ning. He holds a degree in chemical engineering from Punjab Technical University and is a postgraduate in international business management from Indian Institute of Foreign Trade, New Delhi.

Taking the Guesswork Out of CO2 RemovalProTreat® Simulation Software

Get RealDon’t Guess!

SyngaS, Lng and Carbon Capture

ProTreat accurately predicts column performance.

100% Mass-Transfer-Rate based — no ideal stages.

Grassroots? Revamp? Designing for Packing? Using MDEA with piperazine? HotPot?

Build a virtual plant with the industry’s most advanced gas treating simulator.

www.ogtrt.com +1 281 970 2700 [email protected]

Learn how state-of-the-art simulation science can benefit you.

OGT-0117 PTQ Half Horizontal Ad.indd 1 11/13/12 10:40 PM

fluor.indd 3 13/03/2013 12:02

http://www.digitalrefining.com/136,sponsors,Optimized_Gas_Treating.html#.UVAns46kJ94

bete.indd 1 26/02/2013 11:17

http://www.eptq.com/jump.aspx?a=bete&p=gas13

Flame interaction and rollover solutions in ethylene cracking furnaces

Due to an increasing supply of North American shale gas feedstocks, many petrochem-

ical companies are restarting once-idled olefins facilities or plan-ning plant expansions. Since these facilities may have been out of commission for several years, oper-ators should be aware that existing ethylene furnace equipment might not meet today’s emissions require-ments. This article will discuss the challenges facility operators face when retrofitting furnaces with the latest burner technology to achieve required emissions levels.

Retrofitting ethylene furnaces with the latest burner technology can be difficult due to burner spac-ing issues and their resulting effect upon burner flame quality. As existing furnaces are upgraded to achieve higher capacities, more floor burners are added. The addi-tion of more burners can result in smaller distances between burners and more potential for adverse flame conditions, such as burner-to-burner flame interaction and flame rollover.

Flame interaction between burn-ers can cause flame impingement upon the process tubes and increased emissions. Flame impingement can also be problem-atic for prolonged ethylene production, resulting in shorter run lengths and higher tube metal temperatures, which can cause premature coking and lead to shorter periods between decoking, reducing ethylene production and outputs.

One of the other primary concerns for the end user operating company is flame rollover. When

During plant expansion or construction of new facilities for olefins plant, it is important to keep in mind the interactions of flame, fuel and lower emissions

REX K ISAACSZeeco

the momentum of the hot gases moving upwards from the burner becomes less than the momentum of the colder gases moving down the furnace tubes, flame rollover occurs. One product developed to address these concerns is the patented ultra-low NOx GLSF Min-Emissions Enhanced Jet Flat Flame burner. In this article, we will review the design details for this type of burner, provide specific retrofit installation details in a case study format, review lessons learned during the retrofit, and include results for several of the retrofit applications.

An inherent design aspect of the

burner is the fact that the fuel gas is introduced between the furnace wall and the air stream. Consequently, flame interaction between burners is minimised due to the location of the burner tip and the very compact design. Since the gas does not cross the air stream, the tip drilling design can be modi-fied to achieve better heat flux profiles without adversely effecting the thermal NOx emissions.

Application backgroundSince the early 2000s, Zeeco and a petrochemical facility have worked together on many ethylene furnace revamp applications where the number of floor burners increased and the burners had to be moved closer. Even in these situations, using the right technology helps avoid problematic occurrences such as flame rollover and interaction.

To achieve improved flame quality without any flame rollover or flame interaction, the customer selected Zeeco’s GLSF Min- Emissions floor-mounted burners. As Figure 1 shows, this burner is designed to entrain the unburned gas next to the wall to prevent flame rollover as the furnace currents pull the air and products of combustion towards the tubes. As the unburned gas moves up the wall, it mixes with the inert flue gas products of combustion, burn-ing directly below the unburned gas. As the mixture of unburned gas and products of combustion continues to move up the wall, it combines with air and burns. Since the unburned gas is mixed with some of the products of combustion before burning, the peak flame


Furnace currents

Fur

nace

tub

es

Combustion air

Furnace wall

Combustion

Burner

Unburned gas / inert products of

combustion mixture

Inert products of combustion

Figure 1 Computational fluid dynamics (CFD) model of GLSF burner

zeeco.indd 1 13/03/2013 12:17


flux percentage is also decreased. It is important to note that the

location of the staged gas tips also affects the flame quality of the burner. For example, if the burner is required to make very low NOx emissions, the staged orifices must be aimed in a more vertical direc-tion. This vertical direction, coupled with the combustion air stream located in between the unburned gas and the furnace wall, increases the likelihood of flame impinge-ment. The mixing of the fuel gas energy becomes so reduced at higher elevations that the furnace currents can easily influence the flame towards the tubes. In general, the lower the NOx emissions when the staged gas has to pass completely over the combustion air opening, the higher the tendency for flame impingement or hot spots on the tubes.

The latest in low emission burner technology, such as the GLSF Min- Emissions burner, incorporates staged gas tips with gas ports aimed towards the wall in a pattern that provides a very uniform heat flux profile in the middle to upper regions of the flame envelope. The mounting of the burner’s flame-shaping tips on the side of the combustion air

temperature is lowered, producing lower thermal NOx emissions. Therefore, not only are flame rollo-ver and flame interaction problems solved, but ultra-low emissions can be achieved.

Comparison to low emissionburners Low emission burners found in ethylene cracking units typically use staged fuel technology. These particular burners have staged fuel tips strategically positioned for fuel to exit the orifices and pass over the combustion air stream before reaching the furnace wall. In order to modify the flame pattern to achieve an even heat flux in the lower portions of the flame envelope, the orifices must be drilled at increasingly abrupt angles towards the furnace wall. These orifice angles cause the air and fuel gas to mix at a faster rate, thus increasing thermal NOx and requiring a compromise between the heat flux profile and thermal NOx production. As the heat flux profile is made more uniform, with an average above 90%, the NOx emissions typically increase along with the increase in flux percent-age. In the same respect, as the NOx is decreased, the heat

Figure 2 Comparison of a low emissions staged fuel burner and the GLSF burner

Side view of burner

Top view of burner

Low emissions staged fuel burner

Staged gas

Primary gas

Staged gas

Primary gas

Zeeco GLSF Min-Emissions burner

Staged gas

Air flow

Primary gas

Side view of burner

Top view of burner

www.fl exim.com

Bidirectional Measurement

High Dynamic Range

Installation under

Process Conditions

No Risk of Leakage

No Wear and Tear

No Pressure Drop

ATEX-certifi ed

Gas Storage Gas Production Gas Distribution

Non-invasive Gas Flow Measurement

zeeco.indd 2 14/03/2013 10:33

http://www.eptq.com/jump.aspx?a=flexim&p=gas13

stream allows the gas to avoid passing directly over the combus-tion air stream. The primary gas tips and gas lance provide the necessary heat distribution in the lower regions of the fl ame enve-lope. Since the burner mixes internally inert fl ue gases, the fl ame is stretched over a longer distance, enabling a more uniform heat fl ux distribution on the furnace wall. With the fl ame-shap-ing tips located on the side of the combustion air stream, the burner heat fl ux profi le can be changed without adversely affecting NOx or causing fl ame impingement on the tubes. The burner can then evenly transfer heat to the process tubes and reduce the possibility of localised hot spots while produc-ing lower NOx emissions.

Figure 2 shows a comparison between the Min-Emissions burner and a typical staged fuel burner. The typical staged fuel burner, with staged gas tips that normally comprise 70% of the heat release positioned on the back side of the tile, requires that the unburned gas cross the combustion air stream before reaching the furnace wall. The Min-Emissions burner illus-trates that the staged gas does not directly cross the combustion air stream. This type of burner design allows the angle of the staged gas port to be changed without adversely affecting the thermal NOx emissions. Since the volume of air can be around nine times greater than gas, it is very impor-


tant that the gas be injected between the furnace wall and the combustion air stream to keep furnace currents from affecting the fl ame quality.

Small burner size for retrofi t applicationsAnother critical design detail when retrofi tting for emissions control without sacrifi cing fl ame quality is the compactness of the burner

design. Operators should seek a combustion company that engi-neers a burner which is compact in design with no metal located in the throat of the burner, excluding the gas tips. Gas tips are required on each burner to distribute the fuel and mix it with the combustion air stream so that it burns completely. Eliminating any other metal from the burner throat results in a design with fewer items that can fail and require replacement. In addition, by simply eliminating metal from the burner throat, the burner can be

designed with a smaller external dimension. When the external dimensions of the burner are smaller, it can normally replace conventional NOx and staged fuel NOx burners with only minor furnace modifi cations. This compact footprint allowed the petrochemical plant operator to install more burners mounted more closely together in revamped furnace applications without adverse fl ame impact or major modifi cations to the furnace (see Figure 3).

Thermal NOx creation andreductionIn order to understand why the Min-Emissions burner design was successful in this application, the formation of thermal NOx emis-sions must fi rst be examined. For gaseous fuels with no fuel-bound nitrogen (N2), thermal NOx is the primary contributor to overall NOx production. Thermal NOx is produced when fl ame temperatures reach a high enough level to “break” the covalent N2 bond apart, allowing “free” nitrogen atoms to bond with oxygen to form NOx.

A stoichiometric equation describing typical combustion in a natural gas-fi red burner (methane and air with excess air) is as follows:

2CH4 + 4 (XA) O

2 + 15 (XR) N

2 → 2CO

2 +

4H2O + (XA) 15N

2 + (XR) O

2

Natural air is comprised of 21%

GLSF burners mounted side by side

Furnace wall

Figure 3 Left: the GLSF burner in operation with no fl ame rollover or fl ame interaction issues. Right: a diagram depicting the close proximity of the burners in the petrochemical plant revamp application

By simply eliminating metal from the burner throat, the burner can be designed with a smaller external dimension

zeeco.indd 3 13/03/2013 12:18


of the N2 covalent bond decreases, allowing the formation of free nitrogen and subsequently increas-ing thermal NOx. Burner designers can reduce overall NOx emissions by decreasing the peak flame temperature, which can reduce the formation of free nitrogen availa-ble to form thermal NOx.

The varied requirements of refin-ing and petrochemical processes entail the use of numerous types and configurations of burners. The method utilised to lower NOx emissions can differ by application. However, thermal NOx reduction is generally achieved by delaying the rate of combustion. Since the combustion process is a reaction between oxygen and fuel, the objec-tive of delayed combustion is to reduce the rate at which the fuel and oxygen mix and burn. The faster the oxygen and the fuel gas mix, the faster the rate of combus-tion and the higher the peak flame temperature.

Figure 4 plots peak flame temper-ature against thermal NOx created. NOx emissions increase as the adia-batic flame temperature increases. Slowing the combustion reaction reduces the flame temperature, which results in lower thermal NOx emissions. The challenge in achieving lower thermal NOx emis-sions is not the theory; however, the challenge is in retrofitting the latest burner technologies into older existing furnaces without adding expensive external compo-nents or processes.

The industry’s standard method to reduce thermal NOx is to mix the fuel gas together with the inert products of combustion to recondi-tion the fuel before combustion occurs. Since the reconditioned fuel is mainly comprised of inert components, the resulting composi-tion burns at a lower peak temperature. To best utilise the inert products of combustion (flue gas) within the furnace, the fuel gas is introduced along the outside perimeter of the burner tile in an area where flue gas is present while the furnace is in operation. As the fuel gas passes through the inert products of combustion, mixing naturally occurs. This changes the

the free nitrogen is available to bond with other atoms. Basic chemistry dictates that free nitro-gen, or nitrogen radicals, will react with other atoms or molecules that can accept them to create a more stable atom. Of the possible reac-tions with the products of combustion, free nitrogen will most likely bond with other free nitrogen to form N2. However, if another free nitrogen atom is not available, the free nitrogen will react with the oxygen atoms to form thermal NOx. As the flame temperature increases, the stability

O2 and 79% N2. Combustion occurs when O2 reacts and combines with fuel (typically hydrocarbon). However, the temperature of combustion is not normally high enough to break all of the N2 bonds, so a majority of nitrogen in the air stream passes through the combustion process and remains diatomic nitrogen (N2) in the inert combustion products. Very little N2 is able to reach high enough temperatures in the high-intensity regions of the flame to break apart and form free nitrogen. Once the covalent nitrogen bond is broken,

Summary details Number of burners 2 furnaces x 24 per furnace (= 2 rows × 12 per side)Type of burner GLSF Min-Emissions burners with internal fuel gas recirculationType of fuel (gas/oil/dual oil gas) Gas onlyLocation in furnace (roof/floor/side wall) FloorFiring orientation (down-firing/upshot/radiant wall/against wall) Upshot (against wall)Flame shape(round flame/flat flame) Flat flame gas burner assemblyAir supply system (natural/forced/induced/balanced/GTE) Induced draft fan with natural draft burnersMaximum available combustion air pressure at burner, KPa(a) AtmosphericAmbient temperature (minimum/normal/maximum), °C -40.2/+4.5/+36.6Relative humidity, % 70% at 4.5°CAmbient pressure, KPa(a) 99.3Altitude above sea level, m Not applicableAtomisation (mechanical/steam/air) Not applicableFirebox dataFlue gas temperature at cross-over, °C 1143Average flue gas temperature in firebox, °C 1252Firebox volume, m3 568Firebox dimensions (L× W × H), m 2 boxes, each 10.21 x 2.8 x 9.93

Summary information for bottom burners

Table 1

Figure 4 Calculated peak flame temperature vs thermal NOx production

2500

2600

2700

2800

2900

3000

3100

3200

3300

3400

3500

3600

3700

Norm

alis

ed t

herm

al N

Ox

pro

ducti

on

Adiabatic flame temperature, ºF

Porem voluptati quia volorion nos asin corenis tesciis rem de volorio nserrum laut et hiliquam il molores tescium cum

zeeco.indd 4 13/03/2013 12:18


burners located on each side of the tubes.

The radiant wall burners, or “trim” burners, were located in the upper regions of the furnace wall. Each radiant wall burner was mounted in the furnace wall hori-zontally and produced a radial flame pattern on the wall. The radiant wall burners are a pre-mixed design, where air and gas mix together before combus-tion occurs.

The hearth burners provide most of the heat, while the radiant wall burners are the remaining heat release. In order for the heater to work properly, the hearth burners must provide an even heat flux profile for where the tubes are located.

ConclusionAs petrochemical companies continue to revisit their olefins facilities through plant expansion or construction of new facilities, it is important to keep in mind the complex interactions of flame, fuel and lower emissions. Zeeco has provided GLSF Min-Emissions burners for ethylene cracking appli-cations for over 200 cracking furnaces worldwide, with NOx emission guarantees as low as 44 ppmv. As described in the example petrochemical plant installation, using a compact design burner technology to achieve challenging emissions levels without sacrificing flame quality is possible through a retrofit, but it does require careful consideration of options and possi-ble pitfalls.

Rex K Isaacs is the Director of Burner Products for Zeeco, Inc, a combustion equipment manufacturer based in Broken Arrow, Oklahoma, USA. He has over 20 years’ experience in the industry and is listed on five patents for combustion equipment. He holds a BS in mechanical engineering from Oklahoma State University.

composition of the fuel, and stabili-sation occurs at the tile exit. Since the reconditioned fuel mixture is 15-50% inert in most cases, the resulting flame burns at a lower peak temperature and generates less thermal NOx.

The mixing of the fuel gas with flue gas prior to combustion is called internal flue gas recirculation (IFGR). When IFGR is too aggres-sive, it can result in an increased blower power usage, decreased burner turndown and increased flame destabilisation. Through min-emissions theory, maximising IFGR while maintaining flame stability and flame length can become a challenge.

Application and resultsThe petrochemical plant chose to retrofit GLSF Min-Emissions burn-ers in the following cracking furnace design. Each heater required 24 hearth (bottom) burn-ers per furnace. Twelve burners were mounted on each side of the tubes located in the centre of the furnace. The floor burners were designed to fire up the furnace wall, which, in turn, radiates heat to be absorbed by the furnace tubes (see Table 1).

The hearth burners were mounted in the floor of the heater and fired up the firebrick wall. Since the burners fired directly onto the firebrick wall, the wall transferred heat to the process tubes that are located in the centre of the cracking heater. The furnace has two rows of tubes and

Our new clean-fuels plant is straining the

auxiliary units!

How can we boost their capacity without major construction?

Read more on this topic atwww.amacs.com

TO MEET CLEAN FUELS requirements for gasoline, a medium-sized Midwestern refi nery added a low-sulfur fuels technology plant. Gasoline throughput was unchanged. However, the sharp increase in sulfur removal required more hydrogen from the hydrogen unit and sent more sulfur gases to the amine treaters and downstream sulfur units. These auxiliary units became bottlenecks, overdriven at the cost of product purity and amine consumption.

On studying the hydrogen, amine, and sulfur units, AMACS found many opportunities for improving separation effi ciency and capacity. The problems were solved without major construction by applying modern technology to mist eliminators, liquid-liquid separators, and tower trays and packing. Results included haze-free product and reduced amine consumption. Now a diesel clean-fuels plant is being added.

Phone: 713-434-0934 • Fax: [email protected] Emergency Service: 1-800-231-0077

Visit our new website atwww.amacs.com

Porem voluptati quia volorion nos asin corenis tesciis rem de volorio nserrum laut et hiliquam il molores tescium cum

Burner designers can reduce overall NOx emissions by decreasing the peak flame temperature, which can reduce the formation of free nitrogen available to form thermal NOx

zeeco.indd 5 13/03/2013 12:18

http://www.digitalrefining.com/56,sponsors,AMACS_Process_Tower_Internals.html#.UVAn-Y6kJ94

global combustion technology

Harsh operating conditions. Challenging waste streams.

Changing emissions regulations. At Zeeco, we understand your

everyday realities. Our thermal oxidisers and combustion systems are

designed to efficiently dispose of sulphur, halogen, or nitrogen-bound

wastes, BTEX vapors or amine vent streams. We know combustion: our

burners, flares and thermal oxidisers operate successfully on nearly every

continent and in every kind of environment — and our global

offices and experts stand behind each one.

Global focus. Local expertise.

burners • flares • thermal oxidisers • aftermarket products & services

Zeeco, Inc. 22151 E. 91st St., Broken Arrow, OK 74014 USA

+1-918-258-8551 [email protected] zeeco.com©Zeeco, Inc. 2013

Zeeco_PTQ_March_TO.indd 1 3/8/13 10:27 AMzeeco.indd 1 12/03/2013 16:36

http://www.eptq.com/jump.aspx?a=zeeco&p=gas13

Steam reforming natural gas containing higher hydrocarbons

Higher hydrocarbons under typical reforming conditions are converted to methane.

The method described in this article solves non-linear equations describ-ing mass balances for carbon, hydrogen, oxygen and higher hydrocarbons, and calculates reformed gas composition for a modified feed to the primary reformer containing essentially methane by Mearn’s method. The procedure can assist both process engineers tackling the problem of estimating output from a primary reformer at the design stage and plant operating personnel. The solution, approach and program-ming technique can vary.

Estimation of the outlet gas composition of a primary reformer operating with a feed of methane- rich natural gas and steam has been presented by Hampson.1,2 The mass balance equations here are solved to converge on equilibrium constants for the reforming and shift reactions. This involves simul-taneously solving non-linear equations relating to the composi-tion variables and equilibrium constants. The basic requirement is the set of primary reactions and values of equilibrium constants for the reforming and shift reactions. Alternatively, the equilibrium composition is obtained by minimi-sation of the Gibbs free energy function.3 This latter method has the added advantage that a knowl-edge of the set of primary reactions and the values of equilibrium constants is not necessary to deter-mine the equilibrium composition.4 In another method, the primary and secondary reformers are

A method is described for estimating reformed gas composition in a primary reformer when the feed is a natural gas containing higher hydrocarbons

VISHWAS DESHPANDE and SUBRATA SAHA Reliance Ports & Terminals Limited

modelled in Aspen as an RGibbs reactor with RK Soave equations of state and solved to obtain equilib-rium product composition.5 In this article, a method is suggested to tackle higher hydrocarbons such as propane, butane and hexane, which may be present in the natural gas feed to the primary reformer. The actual operating data of a primary steam natural gas reformer are used in the demonstration of the method.

Formulation of the problemPrincipal reactions governing the steam methane reforming process are:

CH4 + H

2O ⇔ CO + 3 H

2 …. Reforming

(H0 = + 206 kJ/mol ) (1)

298

CO + H2O ⇔ CO

2 + H

2 ….. Shift

(H0 = - 41 kJ/ mol ) (2)

The formation of CO2 in the reforming reaction could be consid-ered as an alternative to CO:

CH4 + H

2O ⇔ CO + 3 H

2 (3)

However, only two out of three equations are necessary to repre-sent the overall equilibria. As a general observation, when reform-ing higher saturated hydrocarbons and provided that contact time is long enough, the exit gas composi-tion is that which approximately corresponds to the chemical equi-libria involving the methane-steam and water gas shift reactions (Equations 1 and 2).

Mearns6 has suggested that, for reforming of higher hydrocarbons such as those present in light

naphtha, it has been observed that under typical reforming conditions the only hydrocarbon present is methane. And it is further suggested that the higher hydrocar-bons undergo a complete reaction with water according to the follow-ing equations:

CmH

2n + m H

2O → m CO + (m + n) H

2 (4)

C

mH

2n + 2m H

2O → m CO

2 + (2m +n ) H

2 (5)

The steam reforming of higher hydrocarbons will then be described by either Equation 4 or 5 together with Equations 1 and 2. Thus, higher hydrocarbons present in the feed will undergo reforming following Equations 1 through 5, and at the end of reaction the only gases present will be CH4, CO, CO2, H2 and H2O. These components are then added to the original feed, which contains lower hydrocar-bons. The feed will now contain only lower hydrocarbons — namely, methane — which is the simplest hydrocarbon to reform. It is considered that equilibrium is reached in the shift reaction, but it does not for the methane steam reaction. An allowance is made for this deviation from equilibrium by assuming an approach to equilib-rium, and thus both K1 and K2 values are known at the given outlet temperature.

Method of solutionThe operating data of a primary reformer with natural gas steam feed are shown overleaf. The detailed composition of natural gas to the primary reformer and equilibrium constants for the


reliance.indd 1 13/03/2013 12:23


nH2O = a – 2m + X + Y (7)

nCO2 = m – X (8)

nH2 = 2m + n – X – 3 Y (9)

nCH4 = Y (10)

Total ( ∑ n ) = a + m + n – 2 Y (11)

Equations for equilibrium constants K1 and K2 for reforming and shift reactions can be written as follows:

K1 =

(X - Y) (2m + n - X - 3Y)3 * P

42

(Y) (a - 2m + X + Y) a + m + n - 2Y (12)

K2 =

(m - X) (2m + n - X - 3Y)3

(X - Y) (a - 2m + X + Y) (13)

Kmoles of reformed gases nH2O, nCO2, nH2, nCH4, nCO at equilibrium from the reforming of higher hydrocarbons are obtained by solv-ing Equations 12 and 13, describing equilibrium constants for individ-ual components. These equations are highly non-linear and are solved for conversions X and Y.

Reforming

Ethane 60.59 = (X - Y) (7 - X - 3Y)3 (28.85)2

(Y) (X + Y) (9 - 2 Y)2

Shift

1.229 = (2 - X) (7 -X - 3Y) (14)

(X - Y) (X + Y)

ReformingPropane

60.59 = (X - Y) (10 - X - 3Y)3 (28.85)2

(Y) (X + Y) (13 - 2 Y)2

Shift

1.229 = (3 - X) (10 - X - 3Y) (15)

(X - Y) (X + Y)

Reformingiso and

60.59 = (X - Y) (13 - X - 3Y)3 (28.85)2

n-butane (Y) (X + Y) (17 - 2 Y)2

Shift

1.229 = (4 - X) (13 - X - 3Y)

(X - Y) (X + Y) (16)

Reformingiso and

60.59 = (X - Y) (16 - X - 3Y)3 (28.85)2

n-pentane (Y) (X + Y) (21 - 2 Y)2

Shift

1.229 = (5 - X) (16 - X - 3Y)

(X - Y) (X + Y) (17)

reforming and shift reactions are shown in Table 1.

Operating data of a primary reformer are:• Gasflow=19800nM3/hr • Steamflow=73.2T/hr• Inletgastemperature=830°K• Outletgastemperature=1031°K• Pressureatinlet=3137.28kPa• Pressuredrop=310.66kPa• K2(shiftreaction)=1.229

For estimation of the steam-to-carbon ratio in the feed gas:• Volumetricflowrateofnaturalgas=19800nM3/hr. Considering that1gmole@NTP=22.7Lit• Grammoles/hofnaturalgas= 19800x1000/22.4=883.929Kmoles/hr• Steam=73.2x1000/18=4066.6667Kmoles/hr19800x1000/22.4=883.929Kmoles/hr• Initial

fSteam

pratio=4066.67=

Carbon1039.93753.9104

The complete composition of the feed to the primary reformer is shown in Table 2.

Reforming of higher hydrocarbons For the sake of convenience, thenotations and symbols used in the earlier work of Mearns6 are retained.

For higher hydrocarbons, assum-ing “a” to be the ratio of moles of H2O to moles of CmH2n, then at equilibrium:

CmH

2n + 2m H

2O → m CO

2 + ( 2m + n ) H

2

No of moles a – 2m + X + Y m – Y 2m + n – X – 3Y

CO2 + H

2 ⇔ CO + H

2O

No of moles m – X 2m + n – X – 3Y X – Y a – 2m + X + Y CO + 3H

2 ⇔ CH

4 + H

2O

No of moles X – Y 2m + n – X – 3Y Y a – 2m + X + Y

Equations for calculating Kmoles of reformed gases nH2O, nCO2, nH2, nCH4, nCO at equilibrium from the reforming of higher hydrocarbons are:

nCO = X – Y (6)

CH4 C

2H

6 C

3H

8 i-C

4H

10 n-C

4H

10 n-C

4H

10 i-C

5H

12 n-C

5H

12 CO

2 Air as N

2 Moisture

84.7 8.59 3.39 0.33 0.37 0.06 0.06 0.05 2.14 0.28 0.05

Inlet gas composition to primary reformer, %m

Temperature, °C 758 753 748 743 738K

1 (reforming) 60.59 53.29 46.81 41.07 35.99

Equilibrium constants

Table 1

Species % mole Kmoles/hr Kg/hr Kmoles C Kmoles H2 Kmoles O

2 Koles inerts

CH4 84.7 748.2398 11971.8368 748.2398 1496.4796 0 0

C2H

6 8.59 75.8840 2276.5218 151.7681 227.65218 0 0

C3H

8 3.39 29.9472 1317.6794 89.8417 119.78904 0 0

i-C4H

10 0.33 2.9152 169.0827 11.6608 14.5761 0 0

n-C4H

10 0.37 3.2685 189.5776 13.0743 16.3429 0 0

i-C5H

12 0.06 0.5300 38.1600 2.6500 3.18 0 0

n-C5H

12 0.05 0.4417 31.8024 2.2085 2.6502 0 0

C6H

14 0.03 0.2650 22.7917 1.5901 1.85514 0 0

CO2 2.14 18.9040 831.7760 18.9040 0 18.904 0

N2 0.28 2.4735 69.2585 0 0 0 2.47352

H2O 0.05 0.4417 7.9506 0 0.4417 0.22085 0

Total 99.99 883.31086 16926.4377 1039.9375 1882.9668 19.1248 2.4735

Feed composition to primary reformer

Table 2

reliance.indd 2 13/03/2013 13:44

ReformingHexane+

60.59 =

(X - Y) (19 - X - 3Y)3 (28.85)2

(Y) (96.8281 + X + Y) (25 - 2 Y)2

Shift

1.229 = (6 - X) (19 - X - 3Y)

(X - Y) (X + Y) (18)

Solver functionality in Excel has been used for solving Equations 14 to 18. Converged values of conver-sions “X” and “Y” are for ethane: 1.4542, 0.9756; propane: 2.1319, 1.3912; iso and n-butane: 2.8086, 1.8061; iso and n pentane: 3.4859, 2.2206; and hexane: 4.71613, 2.6356, respectively. Using values of “X” and “Y”, Equations 6 to 11 are solved for higher hydrocarbon species to obtain nH2O, nCO2, nH2, nCH4, nCO at equilibrium in the system. The resultant moles of nH2O, nCO2, nH2, nCH4, nCO are added to check the mass balance. These moles are added to the lower hydrocarbon moles to obtain a final gas composi-tion containing only methane as a component of natural gas. Reformed gas composition for methane is readily calculated as follows.


Reforming of lower hydrocarbons It is assumed that the steam-to-methane ratio in the feed is high enough so that carbon deposition is avoided, and that the equilibrium composition after reforming will depend on temperature, operating pressure and steam-to-carbon ratio.

For a homogeneous system single gas phase reaction:

K1 =

p3H2pCO = n3H

2nCO

fp p

2 (19) pCH4

pH2O nCH4 nH2O ∑n

K2 =

pCO2 pH2 = nCO2

nH2 (20)

pCO pH2O nCO nH2O

For a specified total pressure of 2827.29 Pa and knowing K1 and K2 at given temperatures, the number of moles of each component at equilib-rium is calculated as follows (assuming that “a” is the initial moles of steam per mole of methane):

X = Conversion of methane by Reaction (1) andY = Conversion of CO by Reaction (2)

CH4 + H

2O ⇔ CO + 3 H

2 Reforming reaction

No of moles 1 – X a – X – Y X – Y 3X + Y

Component %w kmol %m %m (Dry)CO 16.8957 107.1754 8.0263 11.5569H

2 7.6708 681.2200 51.0162 73.4574

CH4 0.0643 0.7143 0.0534 0.0770

CO2 33.6376 135.7837 10.1687 14.6418

H2O 41.3414 407.9330 30.5499

N2 0.3899 2.4735 0.1852 0.2667

Reformed gas final composition

Table 3

test run’s results, it could be concluded that the revamp targets for the CDU-1 main fractionator (C-150) were achieved. No hydrau-lic constraint was experienced in achieving the design intake of 13 000 t/d and the required prod-uct quality was achieved.

ConclusionsThe performance of Shell ConSep trays in the HGO pumparound section of the CDU-1 main fractiona-tor met the target of capacity enhancement without any drawback compared to the pre-revamp condi-tions. During the test run, the trays were operating at 10-15% lower than the design capacity even at the design intake of 13 000 t/d due to heavier crude feed and lower feed temperature. However, the built-in capacity margin enabled stable oper-ation for the trays at much above the capacity limit of the first genera-tion of high-capacity trays.

The options to debottleneck columns already equipped with the first generation of high-capacity trays are limited. ConSep trays provide an attractive solution for

www.eptq.com PTQ Q1 2013 77

such cases. In this revamp project, use of only three of these trays in the most capacity-constrained section of the column made it possi-ble to retrofit the existing column and made the capex option more attractive over the other debottle-necking options.

* Shell ConSep, Shell CS and Shell HiFi are Shell trademarks. ** Mellapak Plus 252Y is a Sulzer Chemtech trademark.

References1 Refinery expansion means NZ more self reliant, media release by NZRC, 16 July 2010.2 Wilkinson P M, De Villiers W E, Mosca G, Tonon L, Achieve challenging targets in propylene yield using ultra system fractionation trays, ERTC 2006.

3 De Villiers W E, Bravo J L, Wilkinson P M, Summers D R, Further advances in light hydrocarbon fractionation, PTQ Q3 2004.

KaushikMajumder is Distillation Team Lead of Shell Projects & Technology in Bangalore, India. He holds a bachelor’s degree from Jadavpur University, India, and a master’s and doctorate from Indian Institute of Technology, Delhi. Email: [email protected] Mosca is the Global Refinery Technology Manager of Sulzer Chemtech. He holds BS and MS degrees in chemical engineering from the University “La Sapienza” Rome, Italy. Email: [email protected] is a Process Engineer at Refining NZ. He was the Senior Process Engineer and Commissioning Process Engineer during the Point Forward Project. Email: [email protected]

Parameters Design TestrunFroth backup/CS height, % 68 60Tray pressure drop, mbar 12.3 9.2Tube flood , % 73 60Flow parameter 0.17 0.19Overall column load factor, m/s 0.12 0.10Flooding (CS tray), % 133 112

KeyperformanceindicatorsforConSeptrays

Table3

sulzer.indd 5 11/12/12 18:05:58reliance.indd 3 13/03/2013 12:24

http://www.eptq.com/jump.aspx?a=ohl&p=gas13


References1 HampsonGM,Designprocedures:Asimplesolutiontosteamreformingequations(part1),The Chemical Engineer,523,July1979.2 HampsonGM,Designprocedures:Asimplesolutiontosteamreformingequations(part2),The Chemical Engineer, 621,Aug/Sept1979.3 Smith J M, Van Ness H C, Abbott MM, Introduction to Chemical Engineering Thermodynamics, 6th ed, Tata McGraw Hill,Boston,2001,484-495.4 Rao Y V C, Chemical Engineering Thermodynamics,504-514,1997.5 StraitM,AllumG,GidwaniN,Synthesis gas reformers, www.owlnet.rice.edu/~ceng403/nh3ref97.html(accessedDec2012).6 MearnsA M, Chemical Engineering Process Analysis, Oliver&Boyd,94-120,1973.7 Rostrup-Nielson J,HansenL JC, Concepts in Syngas, Chapter 1 Routes to syngas, 2011,www.worldscientific.com/doi/suppl/10.1142/p717/suppl_file/p717_chap01.pdf(accessedinAug2012).8 Elliott J R, Lira C T, A Workbook for Chemical Reaction Equilibria, 2002, www.egr.msu.edu/~lira/supp/chap14workbook.pdf(accessedinAug2012).9 Keffer D, Chemical Equilibrium, 2001,http://utkstair.org/clausius/docs/che330/pdf/chemical_eq.pdf(accessedinAug2012).10Gordan S, McBride B J, Computer Program for Calculation of Complex Chemical Equilibrium Compositions and Applications, NASA Reference Publication, Part I, Analysis,Report No NASA RP-1311, 1994, www.grc.nasa.gov/WWW/CEAWeb/RP-1311.pdf11Nob Hill Publishing, LLC, Review of Chemical Equilibrium — Introduction, 2011,http://jbrwww.che.wisc.edu/home/jbraw/chemreacfun/ch3/slides-thermo-2up.pdf12Price R M, http://facstaff.cbu.edu/rprice/lectures/multrxn.html(accessedinAug2012)

Vishwas V Deshpande is the ProjectEngineering Manager and Head of theEngineering Coordination & Projects ControlDepartmentwithRPTL(EngineeringDivision),the engineering wing of Reliance Jamnagarrefinery, India. He has wide experience ofthe process and hydrocarbon industries, andholdsaBTechdegreeinchemicalengineeringfrom Nagpur University and MTech inchemicalengineering from Indian InstituteofTechnology,Kanpur,India.Email: Vishwas.Deshpande@ril.comSubrataSahaisHeadofthePipingDepartmentwith RPTL (Engineering Division), theengineeringwingofRelianceJamnagarrefinery,India. He has wide experience in engineeringdesigninthepower(conventionalandnuclear)andhydrocarbonindustries,andholdsaBTechdegreeinmechanicalengineeringfromIndianInstitute ofTechnology, Kharagpur, India, anda PhD from Indian Institute of Technology,Kanpur,India.

Email: [email protected]

such as NLIN in SAS, nmle in R, Optimisation package in Matlab, Mathcad and so on.

ConclusionA method has been developed for the estimation of outlet gas compo-sition for an SMR process where the feed natural gas contains higher hydrocarbons. This could be of use to process engineers at the design stage, when output from the primary reformer is to be esti-mated, as well as to plant operating engineers. The gas composition at the outlet of the reformer is strongly influenced by process conditions. Using this method, it is possible to study the effect of vari-ous important variables such as

hydrocarbon feed characteristics, inlet steam-to-carbon ratio, outlet temperature and outlet pressure, arriving at the optimum operating parameters of the primary reformer.

NomenclatureK

1 Equilibrium constant for steam-methane

reaction=p3H2pCO

pCH4

pH2O

K

2Equilibrium constant for water gas shift

reaction=pCO2pH

2

pCOpH2O

P Reformerpressure,Pa.sX ExtentofconversionforreactionY Extentofconversionforreactiona MolesofH

2O

MoleofCmH

2n

m Carbonatomsinhigherhydrocarbonsn Hydrogenatomsinhigherhydrocarbons

nH2O

,n

CO2,n

H2,n

CH4,n

COKmolesofH

2O,CO

2,

H2,CH

4andCO

PCO

,P

H2,P

CH4,P

H2OPartialpressuresofCO,H

2,

CH4andH

2O

CO+H2O⇔CO

2+H

2Shiftreaction

No.ofmolesX–Ya–X–Y 3X+Y Y

nCO=X–Y (21)

nH2O=a–X-Y (22)

nCO2=Y (23)

nH2=3X+Y (24)

nCH4=1-X (25)

Total(∑n)=1+a+2X(26)

Equations 19 and 20, after substi-tuting the values of moles of individual components from equa-tions 21 to 26, take the following form:

60.59=(X-Y)(3X+Y)3

*28.85

42

(1-X)(4.502-X-Y)5.502+2X(27)

1.229=(Y)(3X+Y) (X-Y(4.502-X-Y)(28)

Equations 27 and 28 are solved using solver functionality, and converged values of X and Y are 0.9947 and 0.5104. Equilibrium moles of gases are nCO, nCO2, nCH4, nH2O, nH2; 2.9968, 0.5105, 3.4946, 0.0053, 0.4843, respectively. From the material balance, moles of CO, CO2, H2, CH4 and N2 are added to the reformed gas kmoles to arrive at a final composition (see Table 3).

ResultsanddiscussionVariation of the reaction conditions has a significant effect on the prod-uct composition, which necessitates close action by process engineers to specify the correct choice of reac-tion conditions. Operation under higher temperatures and lower pressures and high steam-to-meth-ane ratios favours hydrogen gas production; basically, the methane content should be low. For using reformed gas as a fuel gas, the composition should be methane rich and the favourable conditions include lower temperatures, high pressures and lower steam-to- methane ratios (higher than ther-modynamic minimum ratios). The non-linear equations can also be solved using available methods

Variationofthereactionconditionshasasignificanteffectontheproductcomposition

reliance.indd 4 13/03/2013 12:24

The Moscow Refining, Gas & Petrochemicals Week is one of the strongest series of

technology events dedicated to the Russia & CIS Downstream Oil & Gas industry.

Traditionally the 3 conferences gather together more than 500 regional and international

decision makers, attracts up to 50 exhibitors and over 70 high level speakers.

We would like to invite speakers interested in presenting a paper at any of the 3

conferences to send us an abstract of their presentation at the e-mail below.

Case studies and presentations of End-users will be treated with priority.

You are also welcome to contribute to the programme with ideas for topics, format or

speaker suggestions.

Early confirmed sponsors

Sponsoring & Exhibiting

For further information please contact us at:

www.europetro.com + 359 2 859 41 57 [email protected]

RRTC13th Russia & CIS Refining Technology Conference

IGTC5th International Gas Technology Conference

RPTC12th Russia & CIS PetrochemicalsTechnology Conference

MOSCOW REFINING,GAS & PETROCHEMICALS WEEK

16 - 20 September, Lotte Hotel, Moscow

Supported by the Associationof Russian Refiners and Petrochemical Producers

If your company wishes to

sponsor or exhibit at any

of the 3 conferences,

please contact us for more

details. Special rates

available for sponsoring

and/or exhibiting at more

than one event. Exhibition

space is booking very

quickly and we strongly

advise that you confirm

your stand 5 to 3 months

prior to the event.

CALL FOR PAPERS

epc.indd 1 12/03/2013 14:55

Simplify sulfur recovery

and cut your costs

Single source ISO 9001 technology and service provider

www.axens.net

Beijing +86 10 85 27 57 53 Houston +1 713 840 11 33

Moscow +7 495 933 65 73 Paris +33 1 47 14 25 14 Tokyo +81 335 854 985

Low-temperature tail-gas hydrogenation catalyststhat deliver superior and cleaner performance from simplified operations, and significantly lower CO2 emissions. Axens’ TG catalysts can work with steam reheating technologies to reduce energy consumption. Operating costs and investments are also reduced. It’s a winner every time.

axens.indd 1 12/03/2013 14:33

http://www.digitalrefining.com/57,sponsors,Axens.html#.UVAoQI6kJ94

gas cov copy 5 - DigitalRefining | Refining, Gas and ... your raw natural gas contain hydrogen...

Documents

Transcript of gas cov copy 5 - DigitalRefining | Refining, Gas and ... your raw natural gas contain hydrogen...