Demonstration of Essential Reliability Services by a 300 ...Demonstration of Essential Reliability...
Transcript of Demonstration of Essential Reliability Services by a 300 ...Demonstration of Essential Reliability...
Demonstration of Essential Reliability Services by a 300-MW Photovoltaic Power Plant
Presenters:
Clyde Loutan, California ISO (CAISO)
Mahesh Morjaria, First Solar
Vahan Gevorgian, National Renewable Energy Laboratory (NREL)
April 27, 2017
2
California ISO By the Numbers
71,740 megawatts (MW) of power plant capacity (installed
capacity)
50,270 MW record peak demand (July 24, 2006)
27,488 market transactions per day (2015)
25,685 circuit-miles of transmission lines
30 million people served
240 million megawatt-hours of electricity delivered annually (2015)
As of March 2017
3
Resource Mix as of February, 2017
4
California Energy and Environmental Policies
Driving Renewable Integration and Transmission Needs:
• Greenhouse gas reductions to 1990 levels by 2020
• 33% of load served by renewable generation by 2020
• Ban on use of once-through cooling in coastal power plants
• Less predictable load patterns: rooftop solar, electric vehicles, and smart grid
• Over 1,300 MW of electricity storage resources deployed by 2024
• 1.5 million electric vehicles on the road by 2025
• Governor Brown’s 2030 goals
o 50% of the load served from renewables
o 50% reduction in petroleum use – cars & trucks
o 12,000 MW of distributed generation
o Double energy efficiency of existing buildings
o Greenhouse gas reductions to at least 40% below 1990 levels
5
Expected Behind-the-Meter Photovoltaic (PV) Build-Out Through 2020
0
2,000
4,000
6,000
8,000
10,000
12,000
14,000
2016 2017 2018 2019 2020
Behind-the-meter Build-out through 2020
6
Lo
ad
& N
et
Lo
ad
(M
W)
0
1,000
2,000
3,000
4,000
5,000
6,000
7,000
8,000
9,000
14,000
16,000
18,000
20,000
22,000
24,000
26,000
28,000
30,000
32,000
34,000
Load, Wind & Solar Profiles --- Base Scenario
Typical Spring Day 2020
Net_Load Load Wind Total Solar
Win
d &
So
lar
(MW
)
6,700 MW in
3-hours
7,000 MW in
3-hours
13,000 MW
in 3-hours
Renewable Integration: The need for intra-hour & multi-hour flexible capacity
Typical Spring Day
Net Load 9,187 MW
on April 23, 2017
Actual 3-hour ramp
12,960 MW on
December 18, 2016
7
Solar Production Varies from One Day to the Next: First week of March, 2014
-500
0
500
1,000
1,500
2,000
2,500
3,000
3,500
4,000
4,500
6 7 8 9 10 11 12 13 14 15 16 17 18
Day_1 Day_2 Day_3 Day_4 Day_5 Day_6 Day_7 Average
Average
MW
8
Wind Production Varies from One Day to the Next: First week of March, 2014
0
500
1,000
1,500
2,000
2,500
3,000
3,500
0 1 2 3 3 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Day_1 Day_2 Day_3 Day_4 Day_5 Day_6 Day_7 Average
Average
MW
9
16,000
18,000
20,000
22,000
24,000
26,000
-800%
-600%
-400%
-200%
0%
200%
400%
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Ne
t L
oa
d (
MW
)
Ho
url
y C
PS
1
(%)
Hourly CPS1 vs. Net Load --- 01/31/2016
CPS1>=100% CPS1<100% CPS_pass Net Load
ISO Tracks Real-Time Supply and Demand Balance as Measure of Operational Effectiveness ISO supports the Western Grid
when the blue bars are above
100% (red line)
ISO leans on other BAs when the
blue bars are less than 100%
Performance
Target
CPS1 is a statistical measure of a BA’s area control error (ACE) variability in combination with the interconnection frequency error from scheduled frequency. CPS1 assigns each BA a share of the responsibility for controlling the interconnection steady state frequency.
10
Intra-Hour Variability and Uncertainty Could Result in Inability to Control Interconnection Frequency in Real-Time
-1000%
-800%
-600%
-400%
-200%
0%
200%
400%
0
500
1,000
1,500
2,000
2,500
3,000
3,500
Ho
urly
CP
S1
(%
)
Win
d/S
ola
r (M
W)
Wind/Solar vs. CPS1 --- 01/31/2016
CPS1 (Pass) CPS1>=100% Wind Solar CPS1<100%
CPS1 is evaluated on a rolling 12-month average. Over the past few years, the rolling average has been declining as a result of some poor daily performances. Thus, CAISO needs to take measures to improve daily performance on days with higher variability.
11
• Increased requirements for regulation up and down
• Need to manage increased intra-hour flexibility and multiple hour daily ramps
o Approximately 4,000 to 6,000 MW of intra-hour load-following
o Approximately 13,000 MW of continuous up-ramp within a 3-hour time period (almost double current up-ramps)
• Oversupply during midday hours causes increase in curtailment for reliability purposes
• Non-dispatchable resources serving load varies between 8,000 MW to 10,000 MW based on maximum capability of resources
• Opportunities are created for controllable renewable resources to provide essential reliability services
• Opportunities in the California ISO (CAISO) system to leverage capability of new resources have grown at faster rates than previously expected
• Need to comply with a frequency response obligation following a disturbance (Compliance with BAL-003-1)
• Impact of DER resources on the BES is still not fully understood
Summary of Future Grid Operations to Manage a More Complex Grid
12
Meeting Operational Challenges Beyond 33% RPS with Internal and External (EIM) Resources
Generation Storage
Demand Response
Dispatchable
Quick Start
Wider Operating Range (Lower Pmin)
Load Shift
Over Generation Mitigation
Voltage Support
Peak Load Reduction
Off-Peak Load consumption
Dispatchable Wind/Solar
Fast Ramping
Regulation
Frequency Response
13
Goal: Demonstrate the ability of utility-scale PV power plants to provide important ancillary services to the grid.
o Evaluate performance using real field data
o Support CAISO in adapting new operational practices and market mechanisms to integrate higher penetration of renewables
Project:
Team: CAISO, NREL & First Solar
Site: 300-MW PV power plant in California
Testing conducted in August, 2016
Public report released by CAISO in January, 2017
Project Description
http://www.caiso.com/Documents/TestsShowRenewablePlantsCanBalanceLow-CarbonGrid.pdf
14
PV Power Plant Description
• First Solar PV modules
• 4 MVA PV inverters
• 8 x 40 MVA blocks
• 34.5 kV collector system
• Two 170 MVA transformers
• Tie with 230 kV transmission line
• PMUs collecting data on 230 kV side
40 MVA Block
34.5 kV Collection
170 MVA Transformer
230 kV Transmission
PMU 4 MVA Inverter
15
Plant Control System Enables Grid Friendly Features
Power Grid
Transformer
Inverters
PV Module Arrays
Combiner Boxes
Power Conversion Station (PCS) Photovoltaic Combining
Switchgear (PVCS) Substation
• Checks grid’s actual conditions and required set points
• Sends individual instructions to each inverter based on location, losses, and performance
• Controls quality of power coming out of the PV plant
Closed-loop controls at 100 milliseconds!
Operator Enters Set Points
Plant SCADA system
Set Points
Grid Parameters
Patent No. 8,774,974. Real-time photovoltaic power plant control system
16
• Tested remotely from First Solar operations center in Tempe, AZ: o Supervised testing activities
o Tracked plant performance
o Made changes in set points and control parameters as needed.
Testing Process
Tempe, AZ San Bernardino County, CA
Remote Access
17
• FY17 project • Project funding: $140K • Demonstration site:
o 300 MW PV plant in CA
• Project team: o NREL, CAISO, First Solar, plant owner
• Main objectives: o Break new barriers to the utilization
of ancillary services by PV generation o Demonstrate that advanced power
electronics and solar generation can be controlled to contribute to system-wide reliability
o Produce real field data to leverage PV’s value from being simply an intermittent energy resource to providing valuable reliability services
o Dissemination of project results to educate and bridge gaps in perspectives between all kinds of stakeholders
Pioneering Demonstration Project by Funded by SunShot
18
• 20 MW AES Ilumina PV plant • Location: Guayama, PR • Testing conducted during August
2015 • Project team: NREL, AES, PREPA,
GPTech • Controls demonstrated:
o Automatic generation control (AGC) o Primary frequency response (PFR) o Fast frequency response (FFR)
2015 Demonstration Projects in Puerto Rico and Texas
• 20 MW Pecos Barilla PV plant • Location: Stockton, TX • Testing conducted during September
2015 • Project team: NREL, First Solar, ERCOT • Controls demonstrated:
o Ramp control o AGC o PFR o FFR o Voltage control o Reactive power control o Power factor control
Project report: www.nrel.gov/docs/fy16osti/65368.pdf
19
• CAISO-NREL-First Solar custom-developed test scenarios: o Regulation-up and regulation-down, or AGC tests during sunrise, middle of
the day, and sunset o Frequency response tests with 3% and 5% droop settings for over- frequency
and under- frequency conditions o Curtailment and active power control (APC) tests to verify plant performance
to decrease or increase its output while maintaining specific ramp rates o Voltage and reactive power control tests o Voltage control at near zero active power levels (nighttime control).
• More standardized First Solar’s power plant controller (PPC) system commissioning tests: o Automatic manual control of inverters (individual, blocks of inverters, whole
plant) o Active power curtailment control, generation failure and restoration control,
frequency control validation o Automatic voltage regulation at high and low power generation o VAR control o Power factor control o Voltage limit control
Summary of Conducted Tests in CAISO in 2016
20
Testing 300-MW PV Plant
21
Active Power Curtailment Test
22
• 4-sec AGC signal provided to PPC
• 30 MW headroom
• Tests were conducted fat three resource intensity conditions (30 minutes at each condition):
o Sunrise
o Middle of the day
o Sunset
• 1-sec data collected by plant PPC
AGC Participation Tests
Morning
Midday
25
AGC Participation Tests: Continued
Afternoon
26
Typical Regulation-Up Accuracy of CAISO Conventional Generation
AGC Participation Tests: Summary
Time Frame Solar PV Plant Test Results
Sunrise 93.7%
Middle of the day 87.1%
Sunset 87.4%
Combined Cycle
Gas Turbine
Hydro Limited Energy Battery Resource
Pump Storage Turbine
Steam Turbine
Regulation- Up
Accuracy
46.88% 63.08% 46.67% 61.35% 45.31% 40%
Measured Regulation Accuracy by 300-MW PV Plant
27
• 3% and 5% under and over-frequency tests
• 20% headroom
• ±36 mHz dead band
• Actual frequency event time series measured in the U.S. Western Interconnection
Frequency Droop Tests
Droop =∆𝑓/60𝐻𝑧
∆𝑃/𝑃𝑟𝑎𝑡𝑒𝑑
Example of 3% droop test (under-frequency)
Measured droop response
29
Under-Frequency Droop Tests Summary
Up Droop Control Error Statistics (Percentage of Plant Rated Capacity)
Test Type Mean Error (%) Max + Error (%) Max – Error (%) Standard Deviation (%)
3% droop, sunrise 0.21 1.25 -0.34 0.19
3% droop, sunrise 0.17 2.03 -0.09 0.13
3% droop, midday 0.03 1.61 -0.79 0.14
5% droop, midday 0.00 0.95 -0.5 0.1
5% droop, sunset 0.01 0.83 -0.89 0.07
30
Examples of Over-Frequency Droop Tests
Example of 5% droop test (over-frequency)
Measured droop response – 5% Measured droop response – 3%
31
Results of Additional Frequency Validation Tests
32
Comparison of reactive power capability for a synchronous generator and PV inverter:
Reactive Power and Voltage Control Tests
Proposed CAISO reactive capability for asynchronous resources:
33
Measured Reactive Power Capability and Voltages at POI
• Reactive power tests at high and low power production levels
• The plant meets the proposed CAISO reactive power requirements
34
Lagging and Leading Power Factor Control Tests
• ±100 MVAR/min ramp rates applied
• PF limit = ±0.95
• Tests conducted at nearly full power output
35
Reactive Power Control Test
• Demonstrated ability of the plant to maintain capacitive and inductive VARs at the POI
36
Low Generation Reactive Power Control Test
• Plant was curtailed down to 5 MW output level
• Ability of the plant to produce or absorb VARs (±100 MVAR) was demonstrated
• True night VAR support will be demonstrated in future
37
• Smart inverter technology combined with advanced plant controls allow solar PV generation to provide essential reliability services
• Solar PV resources with these advanced grid-friendly capabilities have unique operating characteristics that can enhance system reliability by providing: o Essential reliability services during periods of oversupply o Voltage support when the plant’s output is near zero o Fast frequency response (inertia response time frame) o Frequency response for low as well as high frequency events.
• Accurate estimation of available peak power is important for the precision of AGC control: o It makes sense to include specifications for such available peak
power estimations into future interconnection requirements and resource performance verification procedures
o Perhaps, standardization for reserve estimation methods.
Conclusions
Role of Utility-Scale PV Plants In Grid Stability & Reliability
• NERC identified essential reliability services to integrate higher levels of solar resources
• Utility-Scale PV Plants Provides
Grid Friendly Features Required by NERC Voltage regulation Real power control, ramping, and curtailment Primary frequency regulation Frequency droop response Short circuit duty control Fault ride through
Ancillary Services As Well Frequency Regulation Ramping Capability Flexible Capacity
Utility-Scale PV Plant Contributes to Grid Stability & Reliability Like Conventional Generation
Source: NERC: 2012 Special Assessment Interconnection Requirements for Variable Generation
39
• We’d like to thank the staff of First Solar’s corporate office in Tempe, Arizona for their substantial contribution during the planning and implementations stages of this project.
• We’d also like to thank the CAISO staff who was instrumental in all stages of this project.
• The team also thanks Dr. Guohui Yuan of the U.S. Department of Energy’s Solar Energy Technologies Office for his continuous support of this project.
Acknowledgments
Thank You