Sandia National Laboratories is a multi-mission laboratory managed and operated by National Technology and Engineering Solutions of Sandia, LLC, a wholly owned subsidiary of Honeywell International, Inc., for the U.S. Department of Energy’s National Nuclear Security Administrati on under contract DE-NA0003525 DAN BORNEO – SANDIA NATIONAL LABORATORIES RAY BYRNE – SANDIA NATIONAL LABORATORIES JEREMY TWITCHELL – PACIFIC NORTHWEST NATIONAL LABORATORY January 9, 2019 SAND2018-13308 PE Grid Energy Storage Introductory Training for the New Mexico Renewable Energy Storage Working Group Santa Fe
103
Embed
Grid Energy Storage Introductory Training for the New Mexico … · Energy (kWh) = Voltage (V) difference between anode and cathode multiplied by amount of ion the electrodes are
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Sandia National Laboratories is a multi-mission laboratory managed and operated by National Technology and Engineering Solutions of Sandia, LLC, a wholly
owned subsidiary of Honeywell International, Inc., for the U.S. Department of Energy’s National Nuclear Security Administration under contract DE-NA0003525
DAN BORNEO – SANDIA NATIONAL LABORATORIESRAY BYRNE – SANDIA NATIONAL LABORATORIES
JEREMY TWITCHELL – PACIFIC NORTHWEST NATIONAL LABORATORY
January 9, 2019SAND2018-13308 PE
Grid Energy Storage
Introductory Training for the New Mexico
Renewable Energy Storage Working Group
Santa Fe
2
Agenda
Welcome
Technologies, terms and fundamentals
Demonstrations and case studies
BREAK
Valuation, applications, and resilience
Regulatory proceedings and state policies
Resources
3
Tesla Charging stations Vandalized
3
• The charging stations looked to be pushed back on their pads with one almost off.
• Three of the chargers were missing.
www.richiejeep.com/2018-Cisco-TX-Supercharger/
'Total waste of Creme Eggs': vandals stuff sweets into electric car chargerSOPHIE WILLIAMSThursday 21 June 2018 15:56
▪ Interconnection with other systems (electrical, any non-electrical sources)
▪ Fire protection (detection, suppression, containment, smoke removal, etc.)
▪ Containment of fluids (from the ESS and from incident response)
▪ Signage
37
Improving Storage Safety
Development of
Inherently Safe Cells
• Safer cell chemistries
• Non-flammable electrolytes
• Shutdown separators
• Non-toxic battery materials
• Inherent overcharge protection
Safety Devices and
Systems
• Cell-based safety devices
• current interrupt devices
• positive T coefficient
• Protection circuit module
• Battery management system
• Charging systems designed
Effective Response to
Off-Normal Events
• Suppressants
• Containment
• Advanced monitoring and controls
38
Safety through Codes and Standards
▪ Many ESS safety related issues are identical or similar to those associated
with other technologies
▪ Some safety issues are unique to energy storage in general and others
only to a particular energy storage technology
▪ Current codes and standards provide a basis for documenting and
validating system safety
▪ prescriptively
▪ through alternative methods and materials criteria
▪ Codes and standards are being updated and new ones developed
to address gaps between ESS technology/applications and criteria
needed to foster initial and ongoing safety
39
Introduction to PNNL/SNL Protocol
A set of best practices for characterizing ESS and measuring and reporting their performances
Available at http://www.sandia.gov/ess/publication/
7 Applications
Peak Shaving - Using an ESS to discharge during on-peak periods for electric power while charging the ESS during off-peak periods
Frequency Regulation - Regulate the electric power frequency by providing up regulation by discharging an ESS and providing down regulation by charging
Islanded Microgrids - Using an ESS as an electrical island separated from the utility grid
Renewables Firming (PV, Wind) - Using an ESS to supplement renewable energy generation to provide steady power output
Power Quality - Mitigating voltages sags by injecting real power from ESS for a few seconds
Frequency Control-Using a discharge/charge from an ESS to make up for a sudden loss of generation or load
Model allows estimation of SOC during operation taking
into account
Operating mode
Power
SOC
Temperature
Model has been validated with data
Allows calculation of one way efficiency from rate of
change of SOC
Actual battery performance can be anticipated, thus
providing a high degree of flexibility to the BESS
owner/operator
Self-learning model applicable to energy type of storage
system
Model will be fine-tuned as more data are gathered.
65
Non-Linear Battery Model Enhances Arbitrage Value for SnoPUD
SnoPUD MESA 2UET 2 MW/8 MWh V/V Flow
Annual benefits in energy arbitrage
Key Lesson: Improving operational knowledge
enhances profit potential by finding sweet spots in which
to operate the system to provide services with smaller
profit margins and by minimizing charging losses.
50% more arbitrage revenue possible for
SnoPUD when optimized using self-learning
non-linear battery model
Battery characterization based on data
collected from Avista-operated UET battery
deployed in Pullman, WA.
66
Next Step: Battery State of Health Model
Goal: Develop a reliable and accurate model to predict battery degradation under various
conditions and to integrate it as a module in BSET
Top-down model
Quantifying the effects of energy throughput, charge-discharge power, and operating temperature
Approach being further refined by adding depth of discharge, number of cycles, SOC operation
range, and time at various voltages.
Bottom-up model to estimate battery degradation
The model includes the effect of cycling and calendar aging, taking into account the effect of
temperature and voltage
The model to date accurately predicted degradation after 18 months of testing
Both these approaches will be modified to predict battery degradation across multiple
chemistries – various chemistries within Li-ion and flow batteries.
67
Summary: Valuing Storage Requires a Detailed Methodology
Siting/Sizing Energy Storage
Broad Set of Use Cases
Regional Variation
Utility Structure
Battery Characteristics
Ability to aid in the siting of energy storage systems by
capturing/measuring location-specific benefits
Measure benefits associated with bulk energy, transmission-level,
ancillary service, distribution-level, and customer benefits at sub-
hourly level
Differentiate benefits by region and market structures/rules
Define benefits for different types of utilities (e.g., PUDs, co-ops,
large utilities operating in organized markets, and vertically integrated
investor-owned utilities operating in regulated markets)
Accurately characterize battery performance, including round trip
efficiency rates across varying states of charge and battery
degradation caused by cycling.
68
Energy Storage
Valuation, Applications, and Resilience
69
Energy Storage Applications
Energy storage application time scale
“Energy” applications – slower times scale, large amounts of energy
“Power” applications – faster time scale, real-time control of the electric grid
70
Energy Storage Services (Value Streams)
Source: DOE/EPRI Electricity Storage Handbook in Collaboration with NRECA, 2013J. Eyer and G. Corey, “Energy Storage for the Electricity Grid:Benefits and Market Potential Assessment Guide”http://www.sandia.gov/ess/publications/SAND2010-0815.pdf
71
Why is Storage Valuation Difficult?
Location/Jurisdiction
Market area, e.g., California ISO
Vertically integrated utility, e.g., PNM
Transmission and distribution deferral is very location specific
Many applications require a combination of technical and financial analysis
Dynamic simulations (requires an accurate system model)
Production cost modeling (requires an accurate system model)
Difficult to break out current cost of services, especially for vertically integrated utilities
Identifying alternatives can be difficult
Many storage technologies are not “off-the-shelf ” proven technology (e.g., O&M costs, warranty)
Storage is expensive
72
Storage Valuation Principles
Co-optimization: the system may not fulfill multiple services simultaneously,
and choosing one action may prevent the system from responding effectively to
another opportunity (e.g. discharging for arbitrage may prevent the system from
mitigating an outage)
Performance-informed: asset conditions and performance vary by technology
and design, and we are still learning how precisely systems respond to control
communications and how intensively state of charge (SOC) affects efficiency
Discrete values: benefits must not overlap to avoid double-counting, with a
value developed from an avoided cost, revenue, or societal benefit
Timeframe for analysis: analysis time horizon should be equal to the lifetime
and life-cycle cost of the proposed set of assets
Location: values should reflect local conditions and value streams should be
location-, market-, region-, and utility-specific
73
Energy Storage Value Streams – an example
Renewable firming
For vertically integrated utilities – increased regulating and spinning reserves.
In market areas, adding ramping products.
CA “duck” curve
Solar variability
74
Takeaways
Barriers to energy storage
Cost
Electricity markets/utilities do not properly allocate payments/costs for services
provided
Voltage support
Inertia
Renewable integration
Reliability
The future
Greater penetration of renewables – storage becomes essential
Higher energy prices – storage starts looking better
Efficient market design – helps pay for storage costs
75
Energy Storage Optimization Tool
76
Example: Hourly value at Bainbridge Island
77
Summary of Results (NPV benefits and revenue requirements over 20-
year time horizon) – Bainbridge Island
Random Outages – Mid-C Capacity
Value
Projected Outages – Mid-C Capacity
Value
Projected Outages – Peaker-Driven Capacity Value
Random Outages – Peaker-Driven Capacity Value
Do you notice the biggest contributor?
78
Estimating the Value of Energy Storage – CAISO Example
Analyzed ~2200 LMP nodes in CAISO
Day ahead market arbitrage
Day ahead and real time market arbitrage
Key takeaways
Revenue opportunity is highly location dependent
Significantly more potential revenue if the real time market is included
Storage model1 MW, 4 MWh80% efficiency
79
Results for DA market arbitrage and frequency regulation1
1R. H. Byrne, T. A. Nguyen and R. J. Concepcion, “Opportunities for energy storage in CAISO," accepted for publication inthe 2018 IEEE Power and Energy Society (PES) General Meeting, August 5-9, 2018.
Estimating the Value of Energy Storage – CAISO Example
80
Sterling Municipal Light Department (SMLD)
Sterling Potential value streams:
Energy arbitrage
Reduction in monthly network load (based on monthly peak hour)
Reduction in capacity payments (based on annual peak hour)
Grid resilience
Frequency Regulation
Grid Resilience was the primary goal – other applications help pay for
the system
Several potential value streams (1MW, 1MWh 2017-18 data)
For more information, please refer to:
R. H. Byrne, S. Hamilton, D. R. Borneo, T. Olinsky-Paul, and I. Gyuk, “The value proposition for energy storage at the
Sterling Municipal Light Department,” proceedings of the 2017 IEEE Power and Energy Society General
Meeting, Chicago, IL, July 16-20, 2017, pp. 1-5. DOI: 10.1109/PESGM.2017.8274631
Description Total Percent
Arbitrage $40,738 16.0%
RNS payment $98,707 38.7%
FCM obligation* $115,572 45.3%
Total $255,017 100%
81
Optimization Results – Typical Week SMLD
• Last week of July 2015• Annual and monthly peaks• Spend the majority of the time
at 50% SOC performing frequency regulation
• Charge up to 100% SOC in hour prior to FCM peak
• Discharge for two consecutive hours (FCM and RNS peak)
• Return to 50% SOC and continue performing frequency regulation
• Note minimal arbitrage (qc, qd)• Assumes an energy neutral
(with losses) regulation signal
2 MW, 4 MWh system
82
RNS monthly peak hour
FCM annual peak hour
Discharge for FCM and RNS hours
Get back to ~50% SOC
REG all the time, except RNS, FCM100% SOC
Optimization Results – Typical Day SMLD
83
QuEST – Open Source Energy Storage Valuation Tool
Sandia released the internal Python-based software tools that have been employed in-house since ~2012
Based on Sandia’s Pyomo optimization framework in Python
High level object oriented language for formulating optimization problems
2016 R&D 100 award winner
Open source software package with technical support from Sandia
Initial capabilitiesRevenue optimization in ISO market areas
Arbitrage and frequency regulation
Planned capabilities for subsequent releasesMicrogrid design and operation
Storage plus solar and wind
Technology selection guideTotal cost of ownership calculations using updated data from the Energy Storage Handbook