The Economics of Capacity Payment Mechanisms in

Restructured Electricity Markets

David P. Brown∗

∗ University of Alberta, Department of Economics

Institute for Public Economics

Capacity Market Design 1 / 34

Overview

1 Background

2 Evolution of Electricity Market Design

3 Capacity Market Fundamentals

4 Important Design Details

5 Price Signals and the Energy System

6 Alternatives?

7 Discussion

Capacity Market Design 2 / 34

 Background

Electricity Markets - Background

Traditionally - Vertically Integrated

Market

Concerns over performance

– Rate-of-return regulation
– Investment risk on consumers
– Limited incentives to lower cost
– Economies-of-scale in generation?

Electricity markets restructured

(1980s-Present)

Transmission & Distribution remain

highly regulated

Generation deregulated ⇒ relies on

competitive markets

Capacity Market Design 3 / 34

 Background

Electricity Markets - Background

Restructured markets exist

worldwide
– US, Canada, England, Australia,
Nord Pool, Chile, Brazil, Italy,
Japan (2020), etc.

These complex markets are

coordinated by an Independent
System Operator

Substantial diversity in market

design

Redesigned wholesale power markets

– Principal component - wholesale
procurement auctions

Capacity Market Design 4 / 34

 Background

Electricity Markets - Wholesale Procurement Auctions

Uniform-priced multi-unit auctions

Hourly (or sub-hourly) procurement

auction

Market-clearing price - intersection

of the offer curve and demand

Ideally, the wholesale price is the

primary price signal to drive
investment and production decisions

Electricity markets have numerous

complications:
(i) limited storage + generation Source: MSA (2012)
capacity constraints
(ii) supply=demand must always hold
(iii) market concentration
(iv) highly inelastic demand
Capacity Market Design 5 / 34
 Evolution of Electricity Market Design

Energy-Only Market Designs

Generators must recover all of their fixed cost of capacity investment in energy
markets
– No explicit payment for capacity value

Energy-only markets exist in numerous jurisdictions

– Australia, Nord Pool, Alberta* (prior to 2019), Texas (ERCOT)

Rely on wholesale prices to send right short-run production signals

Rely on scarcity pricing to signal for more investment - prices should rise to reflect
the opportunity cost of a network failure (VOLL)

Real-world complications:
– Concentrated markets → regulations to limit market power (Alberta - OBEGs)
– Trade-off with static and dynamic efficiency (Brown and Olmstead, 2017)
– Price-spikes and scarcity pricing are mitigated by price-caps (e.g., $999.99)
– Demand is highly inelastic → high prices may not reduce scarcity → outages

Capacity Market Design 6 / 34

 Evolution of Electricity Market Design

Energy-Only Markets, Scarcity Pricing, and Missing Money

Scarcity pricing aims to signal the need

for more capacity investment

Price-spikes and scarcity pricing can

have sizable impacts on average prices
(e.g., ERCOT)

Price-spikes can be critical to cover

fixed costs of certain technologies
(Brown and Olmstead, 2017)

Regulators limit price-spikes by

imposing price-caps in wholesale
markets + bid mitigation

Limits scarcity revenues → mitigates

investment incentives
– Problem is particularly acute for
Natural Gas assets

Capacity Market Design 7 / 34

 Evolution of Electricity Market Design

Missing Money, Renewable Support Mechanisms, and

Capacity Markets
Fundamental reliability problem: Highly price-inelastic demand → limited scope for
demand reductions, market power concerns, implementing regulations + price-caps

Other factors can reduce investment incentives in non-renewable capacity (i.e., gas)
(i) Out-of-market renewable subsidy mechanisms → suppress future wholesale prices -
Growing concerns in the EU and US
(ii) Policy uncertainty

Sizable uncertainty over the future energy market revenues for non-renewable
generation in an energy-only market - could be attenuated by allowing more
credible scarcity pricing

Motivated numerous jurisdictions (including Alberta) to adopt capacity payment

mechanisms to “keep the lights on” - aims to correct for the market failure in
investment

Capacity markets compensate resources for the capacity (generation potential)

they provide to the network
Capacity Market Design 8 / 34
 Evolution of Electricity Market Design

Alberta’s Transition to a Capacity Market

Existing framework:
Energy-only restructured electricity market

Limited market power mitigation (OBEGs) - motivate investment

Fossil-fuel heavy portfolio (Gas + Coal 90% of generation in 2015 (AUC, 2017))

Excess generation capacity and reserve margins, low NG prices

Substantial changes in market design and trajectory:

1 Adjusted carbon pricing mechanism - SGER to CCR (Brown, Eckert, Eckert, 2017)

2 Coal unit phase-out by 2030

3 Competitive renewable electricity procurement program

4 30% of generation from renewables

Alberta is choosing to transition to an energy + capacity market design

Capacity Market Design 9 / 34
 Capacity Market Fundamentals

Capacity Market Fundamentals

Transition to a market design with capacity + energy market mechanisms
1 Capacity market fundamentals:
– Compatible with a competitive wholesale market
– Promotes system reliability via more revenue certainty
– Can promote both dynamic and static efficiency

2 Not all capacity market designs are created equal - details matter!
– Learn from experience in other jurisdictions
– Defining details upfront is critical - regulatory certainty (tough with time
constraints)
3 Need to carefully design the market-based energy system
– Price-signals are critical, simply compensating resources for capacity with existing
energy markets may not be enough
– Can lead to misallocation of revenues → inefficient generation portfolio mix
(concerning with growing intermittent renewables)
– Inefficient allocation of capacity costs to load-serving entities can distort incentives
for load-shifting and promotion of demand response

Capacity Market Design 10 / 34

 Capacity Market Fundamentals

Capacity Market Fundamentals

Capacity payment mechanisms are being adopted and considered worldwide

– Adopted: US - (PJM, ISONE, NYISO), Alberta, Brazil, United Kingdom, Chile,
Spain, Peru, among others

Capacity market designs and methodology vary greatly by region

Focus: “Energy Markets with Forward Reserve Requirements and Centralized

Capacity Markets” (e.g., PJM, ISONE, NYISO)
Typical Elements and Timing:
1 Annual capacity market occurs t years in advance of energy markets (e.g., 1 - 5 years)
2 System operator specifies a demand curve for capacity (reserve requirement on LSEs)
3 Generators* submit bids into capacity market (existing and new resources)
4 Uniform capacity price occurs where the bid function intersects demand
5 Load-serving entities are allocated capacity costs based on “coincidental peaks”
6 Resources earn capacity payments (1 - 3 years) - capacity performance obligation
7 Daily and hourly energy markets take place and clear (short-run signals)

Capacity Market Design 11 / 34

 Capacity Market Fundamentals

Uniform-Price Capacity Auctions

EX: PJM’s Reliability Pricing Model - Annual capacity market that occurs 3-years in
advance of delivery-year

Source: Pfeifenberger et al. (2011)

Capacity Market Design 12 / 34
 Capacity Market Fundamentals

Capacity Market Fundamentals

Standard Design Elements and Principals:

Ideally: capacity resources bids’ reflect cost of making capacity available, net of
expected revenues from subsequent energy markets

Generators* can still participate in energy markets if they are not called upon in
capacity markets

Load-Serving Entities (e.g, EPCOR) are required to secure capacity obligations to

meet peak demand
– Can secure capacity obligations bilaterally (outside of the capacity auction)
– The capacity procurement auction serves as a residual capacity market - settles
uncommitted capacity resources and load obligations

The system operator formulates an capacity demand curve that aims to ensure
there is sufficient capacity to meet future demand

Capacity auctions are complex! Success depends critically on the details.

Capacity Market Design 13 / 34

 Important Design Details

Capacity Requirement and Capacity Demand Curve

System operator sets a capacity reliability

requirement - based on peak demand
forecast and reliability criteria (LOLE)

Often downward sloping and based on

numerous administrative parameters

Critical Pivot Points (e.g., PJM, NYISO):

Cost of New Entry of a representative
efficient natural gas combustion
turbine unit
Net of historic energy market revenues
(average of three previous years)
Pivot points are subject to regulator’s Source: PJM (2013)
discretion - impacts slope

Capacity Market Design 14 / 34

 Important Design Details

Capacity Requirement and Capacity Demand Curve

Numerous potential capacity demand curve formulations

Source: FERC (2013)

Capacity Market Design 15 / 34
 Important Design Details

Capacity Requirement and Capacity Demand Curve

Capacity payments 6= stable capacity revenues

Source: PJM (2016)

Capacity Market Design 16 / 34

 Important Design Details

Capacity Requirement and Capacity Demand Curve

Criticisms:

Administrative parameters are controversial and capacity market outcomes very

sensitive to parameters (see plethora of Brattle reports 2009 - Present)
Historical energy market revenues are a poor predictor of future energy revenues
– Important impacts on the capacity demand curve (Pfeifenberger et al. (2011))
– Establish a forward-looking energy + ancillary revenues estimate to formulate
capacity demand curve (Monitoring Analytics, 2016)
Tendency to over procure capacity - need to carefully consider how we treat
interties and renewable capacity values (Newbery, 2016)
Excess procurement can create negative dynamic effects (distorts key signals)
– Suppresses subsequent energy market revenues (creating more “missing money”)
– Increase reliance on capacity market revenues → administrative capacity market
parameters
– Expensive - loss of political support can lead to regulatory policy uncertainty - very
bad for investment

Capacity Market Design 17 / 34

 Important Design Details

Resource Participation
Who can participate?
Traditional fossil-fuel generators (existing and new)

Demand Response - changes in electricity consumption by end-users from their

normal pattern of consumption

– Can yield substantial peak capacity value - called upon to reduce demand during
periods of market scarcity
– Substantial potential here in Alberta (industrial loads)
– Forecast on baseline consumption (Chao, 2011; Brown and Sappington, 2016)
– Capacity and energy market compensation depend critically on estimated baseline
– Concerns over performance in wholesale markets
– Should DR receive capacity payments? (very controversial: suppress capacity prices)
– PJM Lessons Learned: Establish strong performance incentives, requirements, +
penalties for resources participating in capacity markets
2014 “Polar Vortex” - DR (and other) resources receiving capacity payments
not available
Capacity Market Design 18 / 34
 Important Design Details

Resource Participation

Who can participate?

Large penetration of cogeneration - unique to Alberta

– Determining the capacity value (net of behind-the-meter load)
– Should be compensated for providing capacity above this level
– Performance payments for provision beyond this level

Strong performance incentives + penalties are critical to ensure the capacity is

available when needed (lesson learned in the US - PJM’s Capacity Performance
Order FERC 2015)
– Controversial new definition of capacity resources in PJM - strong requirements on
DR and Renewable resources
– Going forward - need a better way to aggregate resources to qualify as an annual
capacity resource (Bothwell and Hobbs, 2016)
– PJM may be willing to act as an aggregator

Capacity Market Design 19 / 34

 Important Design Details

Renewable Generation and Procurement - Alberta

Renewable resources will receive payments via the Renewable Energy Program
procurement auction - 5,000 MWs by 2030
Should they be able to also receive capacity payments?

Great Britain does not allow subsidy supported resources to participate

Some regions include subsidized renewables in capacity markets, but

– Renewables have limited capacity value and strict performance req. (PJM, 2014)

Renewables increase competition in the capacity market, but may suppress

capacity prices because of out-of-market subsidies

Growing concerns in regions that resources that receive out-of-market subsides will
undermine the integrity of the capacity market (e.g., PJM and MOPR - Ohio,
Maryland, Illinois, New Jersey)

If not included, capacity market demand should be reduced by the expected

contribution of renewables to meet peak-load (otherwise, over-procurement
(Bothwell and Hobbs, 2016))

Capacity Market Design 20 / 34

 Important Design Details

Allocating Capacity Market Costs to LSEs

Load-serving entities (LSEs) are required to secure sufficient capacity to meet their
“capacity” obligation (equal to peak-load times a reserve margin)

Some jurisdictions use a “coincidental peaks” forecast methodology to allocate

costs to LSEs (e.g., 5 CPs)
– May not reflect the true relationship between a LSE’s consumers and their need for
generation capacity
– Alternative methodologies rely on previous year’s observed coincidental peaks
– Careful thought needs to go into assessing a LSEs consumers’ contribution to the
capacity requirement of the electric system
– Creates important signals for LSEs to shift-peaks and implement Demand Response
– PJM experience: difficult to change cost allocation mechanisms – critical to get it
right upfront (Monitoring Analytics, 2016)

Capacity Market Design 21 / 34

 Important Design Details

Market Power Mitigation

Alberta’s electricity market is likely to remain relatively concentrated
Market power will be a concern in the energy and capacity markets

Defense for allowing some degree of unilateral market power no longer exists

To mitigate market power, regulator should:

1 Enforce capacity market must-offer provisions for existing facilities (capacity
and energy markets)
2 Create a stable policy framework to encourage investment and participation
in energy and capacity auctions (competition!)
3 Consider imposing bid-mitigation in both auctions on pivotal bidders in
periods of high demand
– Remove OBEGs?
– Three-Pivotal Supplier Test?
– Establish a methodology to compute facility-specific cost estimates (not
technology-specific)
4 Set strong performance incentives and penalties to ensure capacity is
available during “stress periods”
Capacity Market Design 22 / 34
 Price Signals and the Energy System

Price Signals and the Energy System with Increasing

Renewables and Capacity Payments

Incorrect argument: Capacity market + bid mitigation → back to a regulated

market → no more market-based price signals

Reliability is more than just simply investment in MWs - not all capacity is identical

Critical to ensure we are not just throwing $ at MWs

– Misallocation of $ can lead to an inefficient portfolio of assets + costly
– Creates structural market problems poorly suited to meet the needs of the system
– Flexibility is increasingly important with growing renewables

Capacity market operates in conjunction with energy & ancillary services markets
– Designing a robust energy and ancillary market to operate in conjunction is essential
– Market outcomes are critically interrelated
– Capacity markets impact investment ⇒ impact E&AS supply curves
– Expected earnings in E&AS markets impacts capacity market bidding behavior

Capacity Market Design 23 / 34

 Price Signals and the Energy System

Price Signals and the Energy System with Increasing

Renewables and Capacity Payments

1 Wholesale (spot) electricity market

– Compensation for real-time value of electricity production (or demand-reduction)
– Spot market settlements at sub-hourly (e.g., 5 minute) intervals - closely links
prices with current value
– Longer existing settlements dampen the time-varying value of generation -
important with growing intermittent renewables (Alberta: 60 minute average)
– Security constrained economic dispatch - system operator calls upon units given
their spot market bids, ramping up and down capabilities, transmission constraints,
and forecast of renewables (important for real-time reliability with renewables)
– Carbon pricing - implicitly impacts firms’ bids via their cost functions - aims to
internalize the environmental externality of emissions (Brown, Eckert, Eckert, 2017)

Capacity Market Design 24 / 34

 Price Signals and the Energy System

Price Signals and the Energy System with Increasing

Renewables and Capacity Payments
2 Locational wholesale spot market pricing
– Most jurisdictions have adopted some for of location-based pricing
– Accounts for location dependent grid congestion and losses
– More important with growing renewables (far from load centers + heterogenous
geographical value of renewables (Antweiler, 2015))

3 Robust Ancillary Services Markets

– Specialty markets used to ensure supply = demand continuously
– Growing importance with increasing renewables (intermittency + forecast error) -
many jurisdictions increasing the ancillary service products (CAISO, ERCOT)
– EX: Voltage and Frequency control, emergency reserve, operating reserve
– Rewards fast-start, dispatchable, and flexible assets (e.g., natural gas)
– Critical price-signal to flexible resources
– Co-optimized energy + ancillary services markets (SCED) → scarcity pricing penalty
factors during short-term periods of tight supply

Capacity Market Design 25 / 34

 Price Signals and the Energy System

Price Signals and the Energy System with Increasing

Renewables and Capacity Payments

4 Capacity Market (non-renewable technologies)

– Compensates resources for providing capacity value (MWs) to the power system
– Serves as a long-run forward hedging market for capacity
– Not advocating for a technology-specific capacity procurement
– Capacity market price-signal is net of price signals from energy and ancillary services
markets (reward for other technology-specific asset capabilities)
– Elevates the importance of avoiding over procuring capacity in the capacity market
- cause distortion of important price-signals in other markets
5 Renewable Energy Program
– Aims to subsidize renewable generation technologies
– Price-signal for renewable capacity resources into the market
– Contracts-for-differences design implies clearing price is indexed to spot price

Capacity Market Design 26 / 34

 Price Signals and the Energy System

Summarizing Price Signals

Complex web of interactions and price-signals

What is the “right” and most efficient combination of markets to drive the optimal
portfolio of generation units - very complicated
This is the trajectory of the Alberta market → lets ensure that this complex web is
designed to provide efficient price signals for asset value
Critical to “right” drive investment + ensure efficiency of the power system in
Alberta

Price-Signals Summarized
1 Wholesale Market - real-time locational value of electricity (including emissions) + SCED
2 Ancillary Services - real-time grid reliability; supply=demand; asset flexibility; scarcity
pricing
3 Capacity Market - long-run reliability value of capacity (in MWs)
4 Renewable Procurement - long-run investment of renewable capacity (in MWs)

Capacity Market Design 27 / 34

 Alternatives?

Are Capacity Markets the Only Way?

Energy-only markets are viewed as the dominant alternative to capacity markets
Rely only on energy and ancillary service markets to motivate investment
May not be enough to motivate future investment
– “Missing Money” problem (price-caps, inadequate ancillary service compensation,
bid mitigation, etc.)
– Reduced investor confidence because of subsidized renewables and policy
uncertainty
– “Missing Market” Problem - long-run market to hedge investment risks (Newbery,
2016)

Experiences:
1 Success: Texas - Energy-Only market with a lot of Wind (15.4% generation)
2011 - faced capacity shortage concerns + low reserve margins
2014 - Elevated wholesale price-cap to $9,000
New forecasted reserve margin in 2020 (-0.8% to 20.5%)
Why?
– (i) Reduced demand growth
– (ii) Large investment in Wind (and a bit of solar)
– (iii) Large investment of Natural Gas CC (waiting out coal retirements)
Capacity Market Design 28 / 34
 Alternatives?

Are Capacity Markets the Only Way?

2 England chose to recently adopt a capacity market - concerns that subsidized
renewables undermine the economics of private natural gas investment

3 Australia’s National Electricity Market - growing renewables and high capacity

reserve margin (unexpected low demand growth)
– 2016 - 2017 outages (issues with market design + ancillary services and potentially
strategic behavior, not adequacy)
– Growing movement to alleviate market power + redesign the market

4 Nord Pool - High scarcity pricing, successfully attracts sufficient capacity

investment

5 Numerous regions in the United States (e.g., PJM, ISONE, NYISO) adopted a
capacity market years ago because of concerns over the “missing money problem”

6 MISO and numerous other jurisdictions - operate under cost-of-service

(rate-of-return) regulation

7 CAISO - bilateral contracting approach + reserve requirement - opaque and high

transaction costs
Capacity Market Design 29 / 34
 Alternatives?

Are Capacity Markets the Only Way?

What is the “right” market design (and reserve margin)?

Depends critically on the region’s policy objectives and risk
preferences:
– Energy-Only Markets: potentially the most economically efficient
outcome if the anticipated reserve margin, outage probabilities
(reliability), allowance of periodic high price-spikes due to scarcity and
market power are publicly acceptable. Comes at the cost of heightened
resource adequacy risks, price-spikes, + trade-off between static and
dynamic efficiency.
– Capacity Payment Mechanisms: likely most appropriate market
design if primary objective is resource adequacy, hitting reserve margin
targets, and avoiding high price events. Comes at the cost of
potentially higher overall costs, shifting some investment risks onto
consumers, and higher administrative burdens.

Capacity Market Design 30 / 34

 Discussion

Conclusions

“Missing Money” + subsidized renewables + policy uncertainty → creates

concerns of future investment and resource adequacy in energy-only markets

Capacity markets aim to promote investment by providing long-run revenue

certainty (dynamic efficiency)

Compatible with traditional energy-markets that can promote static efficiency

Critical that regulator’s think carefully about the synchronization of the energy
system with growing renewables + capacity markets - price signals remain

All capacity is not created equal - establish well-functioning markets to send the
right price signals - (not suggesting technology-specific capacity payments)

Capacity markets provide more reliability, but are complex and potentially more
expensive than an energy-only market design

Capacity Market Design 31 / 34

 Discussion

Conclusions

Details matter - numerous issues to iron out before implementation

1 Defining capacity requirement and LSEs capacity obligations

2 Capacity value of renewables, DR, EE, interties ⇒ minimize over-procurement

concerns

3 Who receives capacity payments - carefully define “capacity resource” (DR, EE,
cogen, imports, renewables)

4 Performance incentives and penalties (e.g., PJM issues)

5 Allocation of capacity costs to LSEs

6 Treatment of market power in capacity and energy-markets

7 Adjust existing E&AS markets to send efficient price-signals with growing

renewables - impacts investment and capacity market outcomes (do not wait)

Capacity Market Design 32 / 34

 Discussion

Alberta-Specific Policy Thoughts

Tight time-frame (2019!) → careful balance between attempting too much and
doing too little
– Risky to focus only on capacity market, and deal with other market details later

In addition to considering the standard “best-practices” capacity market approach

(e.g., 3-year forward, uniform-price, capacity demand curve, etc.):
– Careful thought process on how to treat Demand Response (participate in capacity
market?) and cogeneration
– Set strong performance requirements + penalties (avoid PJM issues)
– Location-specific wholesale pricing (geographical diversity in generation)
– Security constrained economic dispatch + more granular spot pricing
– Must-offer requirements + market power mitigation in both capacity and energy
markets
– Forward-looking net offset of E&AS in capacity demand model
– Careful treatment of transmission interties and renewable capacity value
– Careful thoughts on how to allocate capacity-related costs to LSEs

Important to establish the operating principals of the energy system going forward
to provide market design certainty
Capacity Market Design 33 / 34
 Discussion

References
Antweiler, W. (2015). “A Two-Part Feed-in-Tariff for Intermittent Electricity Generation.” SSRN Working Paper 2649205.
AUC (2017). Annual Electricity Data Collection. Alberta Utility Commission.
Bothwell, C. and B. Hobbs (2016). “Crediting Renewables in Electricity Capacity Markets: The Effects of Alternative Definitions
upon Market Efficiency.” Working Paper.
Brown, D. and D. Sappington (2016). “On the Optimal Design of Demand Response Policies.” Journal of Regulatory
Economics.
Brown, D. and D. Olmstead (2017). “Measuring Market Power and the Efficiency of Alberta’s Restructured Electricity Market:
An Energy-Only Design.” Canadian Journal of Economics.
Brown, D., Eckert, A., and H. Eckert (2017). “Carbon Pricing with an Output Subsidy under Imperfect Competition: The Case
of Alberta’s Restructured Electricity Market.” SSRN Working Paper Series 2928821.
Chao, H. (2011). “Demand Response in Wholesale Electricity Markets: The Choice of the Consumer Baseline,” Journal of
Regulatory Economics.
FERC (2013). Centralized Capacity Market Design Elements. Commission Staff Report AD13-7-000. Federal Energy Regulatory
Commission.
Monitoring Analytics (2016). Design Considerations for an Alberta Capacity Market.
MSA (2012). “Assessment of Static Efficiency in Alberta’s Energy-Only Electricity Market.” Alberta Market Surveillance
Administrator.
Newbery, D. (2016). “Missing money and missing markets: Reliability, capacity auctions and interconnectors” Energy Policy.
Pfeifenberger, J., Newell, S., Spees, K., Hajos, A., and K. Madjarov (2011). “Second Performance Assessment of PJM’s
Reliability Pricing Model.” The Brattle Group.
PJM (2013). PJM Manual 18: PJM Capacity Market. Prepared by: PJM Capacity Market Operations.
PJM (2014). PJM Renewable Integration Study. Prepared by: General Electric International.
PJM (2016). 2019/2020 RPM Base Residual Auction Results.

Capacity Market Design 34 / 34