Outline of Talk Transmission Investment and Access Valuing - - PowerPoint PPT Presentation

SLIDE 1

1

Transmission Investment and Access

Frank A. Wolak Department of Economics Stanford University Stanford, CA 94305-6072 wolak@zia.stanford.edu http://www.stanford.edu/~wolak Chairman, Market Surveillance Committee

• f the California ISO

2

Outline of Talk

• Valuing Transmission in Vertically Integrated (VI) Regime

– Engineering Reliability

• Valuing Transmission in Wholesale Market (WM) Regime

– Economic Reliability

• Why more transmission capacity is needed in wholesale market regime

– Local market power – Transmission network as facilitator of commerce

• Why very little transmission capacity has been built over past 30 years in US
• Methodology for Valuing Transmission Upgrades in VI Regime
• Value of economically reliable transmission network to California
• Paying for transmission expansion in wholesale market regime

– Economic cost causation principles

• Congestion Revenue Rights (CRRs) under Locational Marginal Pricing (LMP)

– What they can and cannot hedge

• Efficient CRR allocation

– A method for efficient CRR allocation

3

Valuing Transmission in VI Regime

• VI utility has two choices for meeting an increase in demand at a given

location

– Construct local generation – Increase transmission capacity to bring in more distant generation

• VI utility’s retail price of electricity is regulated

– Profit-maximizing response of VI utility is to choose system-wide least-cost solution to meet local load increase

• Value of transmission expansion in VI regime

– Displace high cost local generation with lower cost distant generation – Suppose it costs VI utility $50 MWh to produce energy locally but it can import energy at $20/MWh

• Upgrade of 10 MW of capacity implies benefit of $300 = [($50/MWh -

$20/MWh)*10 MWh] during hour

• Value of transmission is cost-saving to VI from increased ability to exploit

locational cost differences

– Reliability value of upgrade can be handled in this framework

• Upgrade of 10 MW to eliminate 0.01 probability of outage where 10 MWh of

demand is curtailed at a cost of $10,000/MWh

• Expected benefit of upgrade is $998 = [0.01*($10,000/MWh - $20/MWh)*10 MWh]

4

Valuing Transmission in VI Regime

• VI utility has legal requirement to meet all demand at regulated retail price

– This implies that once retail price is set, revenue stream of VI utility does not depend on its production decisions

• VI may find it optimal to meet local energy need with high cost local

generation

– Install 50 MW unit that costs $150/MWh to operate instead of upgrading local transmission network

• Transmission upgrade with low-cost distant generation entails significantly

more regulatory risk

– Requires discrete transmission capacity expansion

• Much longer time horizon to construct transmission versus local generation

– Can require larger distant generation investment to realize economies to scale – Commitments of previous regulator must be honored by current and future regulators

• Lower risk local generation solution creates regions with insufficient

transmission capacity into region to meet all demand with distant generation

– San Francisco Bay Area – San Diego Area (2,300 MW local generation to meet 4,500 MW peak load)

SLIDE 2

2

5

Engineering Reliability

• Enough transmission capacity so that

– Demand at all locations in network can be met with pre- specified probability – Assuming that vast majority of generation units in network are owned and operated by same entity

• Because of structure of regulatory process in VI

regime, strong incentive for VI to operate its generation units to limit congestion

– Utility interested in minimizing total cost of supplying all

• f retail load

– No incentive to operate high cost units more intensively to increase locational price differences

• This only increases total costs of VI utility which reduces its profits
• Recall VI utility’s revenue stream is independent of its actions

6

Differences in Supplier Incentives

• Regime 1—VI utility (that owns local generation and

transmission network) had fixed price contract with retailer in DPL South

– Strong incentive to limit locational price differences

• Regime 2—Fixed price contract with retailer in DPL South

ended

– Strong incentive to increase locational price differences because this increases value of VI utility’s local generation holdings and Congestion Revenue Right (CRR) holdings

• Regime 3—Large retailer divested large amount of DPL South

capacity to merchant generation owner

– Strong incentive to increase locational price differences between both DPL South (merchant supplier) and DPL North (large retailer) and

• ther PJM locations

7

Incentives in Action in PJM Market

Real-Time Prices Average Regime 1: 06/1/98 - 07/22/99

0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 80.00 90.00 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Hour $/MWh

PJM West Hub DPL South DPL North

8

Incentives in Action in PJM Market

Average Real-Time Prices Regime 2: 07/23/99 - 06/24/01

0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 80.00 90.00 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Hour $/MWh PJM West Hub DPL South DPL North

SLIDE 3

3

9

Incentives in Action in PJM Market

Average Real-Time Prices Regime 3: 06/25/01 - 06/20/03

0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 80.00 90.00 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Hour $/MWh PJM West Hub DPL South DPL North

10

Valuing Transmission in WM Regime

• Value of transmission expansion is increased ability to exploit locational price

differences

– Note that in VI regime locational cost differences are source of value of transmission link – In WM regime locational price differences are source of value of transmission link

• Incremental source of value for transmission expansions in WM regime
• Why should locational prices differ significantly from costs?

– Local market power—The ability of local suppliers to raise price and profit from this price increase

• Why does local market power arise?

– Insufficient competition among suppliers because transmission network limits size of market any supplier can compete in

• Transmission network configuration can be used strategically by suppliers to raise the prices

they are paid

– Single generation owner can own all capacity needed to meet local demand net of transmission capacity into region

• Local demand is 100 MWh
• Transfer capacity into region is 60 MW
• Single local supplier owns 100 MW of capacity

– Local supplier is monopolist for 40 MW of energy and can name price this amount of energy or lights go out

11

Valuing Transmission in WM Regime

• Factors that exacerbate local market power problem

– Generation assets typically divested in tight geographic clusters

• While there are efficiency reasons for allowing this geographic pattern of divestiture, regulator

should account for local market power costs of this decision

– Decisions of former VI to meet its total load obligations at least cost led to construction

• f high variable cost units in load centers
• Load center—area with less local generation than load to serve
• Solutions to local market power problems

– Reliability Must-Run units—Annual fixed payment to unit owner in exchange for option

• f ISO to call unit (if it is available) to meet local reliability need and pay regulated

variable cost for energy – Automatic Mitigation Procedure (AMP)—Mitigate bids of units that bid substantially in excess of their reference level and have a significant impact energy costs – Warning—These solutions are imperfect.

• Fixed payments in RMR contracts are substantial

– Approximately $1 per MWh of delivered energy – Payments above market-clearing price

• AMP only prevents substantial exercise of market power

– These solutions do not limit incentive suppliers have to exercise local market power,

• nly limit their ability to do so
• Particularly true for AMP mechanism
• Transmission expansions limit incentive of suppliers to exercise local market power

12

Valuing Transmission in WM Regime

• Transmission capacity as solution to local market power problem

– Expands number of suppliers that can compete to supply energy at any location in transmission network

• Flattens slope of residual demand curve faced by any supplier bidding into wholesale market

– Reduces incentive supplier has to exercise local market power

• Withholding capacity to drive prices up only reduces output sold, but does not increase price
• Transmission upgrades have an additional source of benefits in WM regime

– Limit locational price differences due to the exercise of local market power

• Transmission upgrades are particularly valuable during extreme conditions

– Large inter-connection between WSCC and eastern US during period June 2000 to June 2001 was worth on the order of $30 billion – Had significant inter-connection between eastern US and WSCC been in existence, prices in WSCC would not have risen to levels that existed during period May 2000 to June 2001

• Transmission upgrades have significant option value against extreme events

– In hydroelectric-dominated system such as California value of transmission upgrade even greater because water supply does not respond to price of electricity – Low water year in Pacific Northwest can lead to substantial opportunities for California suppliers to exercise unilateral market power (increases slope of residual demand curve)

• Wolak, Frank A., “Diagnosing the California Electricity Crisis,” The Electricity Journal, August

2003.

SLIDE 4

4

13

Valuing Transmission in WM Regime

• Example of cost differences versus price differences from California

market

– Variable cost = (heat rate)*(gas price) + variable O&M – Difference in variable cost of highest cost units operating SP15 and NP15 during congested periods should be no more than $25/MWh to $20/MWh

• Difference in heat rates times gas price

– 10,000 BTU/kwh unit in SP15 versus 15,000 BTU/kwh unit in NP15 – $5/MMBTU price of gas yields $25/MWh variable cost difference

• Locational price differences across congestion zones where as high as

$750/MWh during period with $750/MWh price cap in ISO’s real-time market

– If supplier possesses substantial local market power this is very likely to occur

• Value of transmission link in WM regime significantly higher because it

limits frequency of high locational price differences due to local market power

– Potential for unilateral exercise market power can make extreme locational price differences for sustained periods of time very likely

• Recall that all or even a significant fraction of local market power cannot

be mitigated away with RMR contracts or an AMP mechanism

14

Transition to WM Regime

• Many loads located in regions that are served by expensive local generation because of

decisions by VI utility

– Current transmission network built to serve VI regime – Generation and transmission network built to serve total demand of VI utility at least cost, not facilitate a competitive wholesale electricity market

• In wholesale market regime consumer pays locational price for all energy consumed, versus

average total cost for utility for VI regime

– Unless they have pre-existing long-term supply contract – Move to wholesale market regime can shift costs onto certain groups of consumers

• Cost shifts occurs because WM operates on transmission network that it was not designed for

– Suppliers receive and consumers pay locational price for energy at each location

• Suppliers with local market power raise locational prices they receive and consumers pay

– Congestion Revenue Rights (CRRs) can help to limit magnitude of costs shifts, but there is always residual congestion risk

• CRRs can enhance incentive of suppliers to exercise of local market power

– California ISO’s proposal to have all demand pay average zonal price limits magnitude of residual congestion risk—but this dulls incentive of retailers to sign forward contracts

• Transmission upgrades necessary in transition to WM regime

– Limit cost shifts due to wholesale market operating on a transmission network it was not designed for – Produce more equitable distribution of residual congestion risk 15

Why Transmission Has Not Been Built

• “Expanding U.S. Transmission Capacity,” by Eric Hirst estimates that between

1975 to 1998 transmission investment in US fell by an average of 115 million 1997 dollar period year

– From $5 billion in 1975 to slightly more than $2.5 billion in 1998

• Hirst also reports that (MW-miles transmission capacity)/(Peak Summer Demand)

declined by more than 10 percent for all but one NERC region over period 1989 to 1998

• Under WM regime transmission is no longer on equal footing with local generation

within same firm to meet local energy need

– Most generation owners financially independent of transmission owner

• May not want transmission upgrade because it reduces ability to exercise local market power

– Transmission owner earns regulated rate of return

• May not want to undertake transmission upgrades if return offered is less than other investment
• pportunities

– Congestion charges are a cost of wholesale power borne by all retailers serving given geographic area

• Cost of doing business in that location in the network for all retailers
• Conclusion—There may be no market participant in WM regime with incentive to

invest

16

Economic Reliability

• Sufficient transmission capacity so that all locations in the

network face significant competition from enough independent suppliers to cause them to bid close to their marginal cost curve the vast majority of hours of the year

– All suppliers face sufficiently elastic residual demand curves a large fraction of hours of the year

• Generation divestiture decisions can increase the economic

reliability of a given transmission network

– Conversely, to the extent that significant generation divestiture cannot be implemented, more transmission investment may be needed to achieve economic reliability

• Transmission network facilitates commerce in same way that

inter-state highway system facilitates commerce US economy

– US Highway system built at a cost of 330 billion 1996 dollars – Net benefits from system vastly in excess of this magnitude

SLIDE 5

5

17

Valuing Transmission Upgrades

• Supplier’s expected profit-maximizing bids depend on

distribution of residual demand curves faced

• Distribution of residual demand curves any supplier faces

depends on transmission network configuration

– With more transmission capacity, faces suppliers with competition from more independent suppliers – Reduces slope of residual demand curve this supplier faces

• In quantifying net benefits of given transmission configuration

must recognizes that

– Changes in network configuration will change expected profit- maximizing bidding behavior of suppliers

• Conclusion--Transmission network configuration choice

should recognize and anticipate expected-profit maximizing responses of suppliers in selecting welfare-maximizing network configuration

18

Valuing Transmission Upgrades

• Similar logic can be applied to entry decisions of suppliers
• Net benefits of transmission configuration recognizes

– How changes in network configuration impact prices at each location because expected- profit-maximizing entry decisions of suppliers change – Transmission configuration choice must recognize and anticipate expected-profit maximizing entry and bidding responses of suppliers in selecting welfare-maximizing network configuration

• Build transmission network in anticipation of expected-profit maximizing response
• f suppliers

– Do not build transmission network to react to entry decisions of suppliers as is the case for FERC’s Large Generation Inter-connection Policy

• Generation owner locates, pays for upgrades and then will be reimbursed by market participants

for cost of upgrade

• Suppliers will locate based on what is privately profit-maximizing for them

– FERC Large Generation Inter-connection Policy approach is likely to lead to extremely inefficient and expensive transmission network relative to transmission network built to meet economic reliability criterion

• No economic cost-benefit test for a given upgrade
• Economic reliability approach would only undertake those upgrades with positive

net benefits

– Net benefits account for expected profit-maximizing entry and bidding decisions in response to change in transmission configuration

19

Build Transmission to Facilitate WM

• Time lag to build transmission facilities substantially longer than lag to build

generation facilities provide further justification for economic reliability approach

• Building transmission in response to generation entry will be a continual process of

catching up with consumers always bearing the cost of catching up

• Building transmission network recognizing that supplier

– Will enter where it is profit-maximizing to do so – They will bid to maximize profits once they enter

• Current cost of transmission network is small part of delivered price of electricity in

California

– Roughly 0.2 cents/KWh delivered is average cost of transmission network for California ISO control area – Average retail price of electricity close to 13 cents/KWh – Undertaking upgrades that double transmission charge would come much closer to economically reliable transmission network for California

• Economically reliable network would substantially increase competitiveness of wholesale

market

– Reduce need for RMR contracts and associated fixed and variable payments – Recall that RMR contract fixed payment roughly equal to 0.1 cent/KWh

• More competitive wholesale market would very likely lead to average wholesale energy price

reductions greater than increase in transmission charge, so that delivered price of electricity would fall and retail prices would fall 20

Value of an Economically Reliable Transmission Network to California

• Pacific Northwest has approximately 30,000 MW of hydroelectric capacity
• The Western US has vast coal reserves that can produce power very

inexpensively

• Enhanced transmission inter-connections with rest of WSCC can increase

amount of low cost power that can sell into California

– California can purchase coal-by-wire from mine-mouth power plants located in Montana and Wyoming – Provide fuel supply diversity to California which is very dependent on natural gas for all existing and proposed generation facilities

• Reduce slope of residual demand curve faced by in-state suppliers

– A major factor in California crisis was reduced import availability from rest of WSCC which increased slope of residual demand curve faced by instate suppliers

• See Wolak, F.A. “Measuring Unilateral Market Power in Wholesale Electricity

Markets: The California Market 1998 to 2000,” American Economic Review, May 2003.

SLIDE 6

6

21

Paying for Transmission Network

• Economic cost causation principles imply that market

participants should pay for transmission upgrades that are caused by their injections and withdrawals from the system

• Very difficult, if not impossible, to determine causally related

costs to assign to specific market participants associated with network upgrades

– As a result of withdrawing 1 MWh from system at certain location in the network, how much more do transmission costs increase, given the level

• f withdrawals of all other market participants
• Incremental cost of this load’s withdrawals
• For the costs of connecting to transmission network to

generation owner

– Incremental cost is equal to total cost

• For network upgrades incremental costs are likely to be

extremely small because all loads must buy power from common transmission network

22

Paying for Transmission Network

• This logic suggests most of costs of transmission upgrades are

common costs that must be allocated among market participants in some “fair” manner

– Similar logic to allocating costs of inter-state highway system—Recall 90-10 Federal-state ratio

• Most of costs are common costs not caused by local firms or households
• This argues in favor of uniform $/MWh charge to fund all

network upgrades

• Transition costs associated with going from network that

satisfies engineering reliability criterion to one that satisfies economic reliability criterion argues for uniform $/MWh charge

• Small share of transmission charges in retail price of electricity

argues in favor of $/MWh charge

– Devising alternate cost allocation methodologies are likely to be extremely contentious and expensive

23

Congestion Revenue Rights

• Congestion Revenue Rights (CRRs) are typically obligations to the difference

between the price at the sink and price at source

– Suppose price at sink is $50/MWh and price at source is $40/MWh

• Owner of 1 MW CRR receives $10 = ($50/MWh - $40/MWh)

– If price at sink is $30/MWh and price at source is is $50/MWh

• Owner of 1 MW CRR pays $20 = ($50/MWh - $30/MWh)
• How CRRs are allocated can impact how generation owners bid into wholesale

market

– If have ability to influence locational prices, can earn additional revenues through CRR payments by raising or lowering prices at certain locations

• Market participant able to produce the most congestion revenues with CRRs in their

possession is the one that will value it most

– This implies that any auction mechanism for allocating CRR will result in the supplier able to cause the most congestion to own the CRR

• This logic implies that CRRs should be allocated to load

– ISO’s CRRs should be allocated to loads and move with load as its switches retailers

• Any market participant can buy or sell a CRR, but seller is essentially creating its
• wn CRR that is not backed by ISO
• Caution--Even transmission owner affiliated with owner of CRRs can even profit

from transmission outages through CRR holdings

24

Using CRRs to Profit From Transmission Outages in PJM Market

• Let TRANOUT(h,d,y) = indicator that is equal to one if there is

a transmission outage in hour h of day d in year y

• PD(DPL Zone, h,d,y) = real-time DPL Zone price in hour h or

day d in year y

• PD(WH,h,d,y)) = real-time Western Hub price in hour h of day

d in year y

• OUTCONG(h,d,y) = TRANOUT(h,d,y)*(PD(DPL Zone,h,d,y)

– PD(WH,h,d,y))

• CCONG(h,d,y) = (PD(DPL Zone,h,d,y) – PD(WH,h,d,y))
• OUTCONG measures congestion due to transmission outages
• CCONG measure congestion due to all causes
• Transmission owner is affliate of load-serving entity (that owns

local generation) that has substantial CRR holdings

SLIDE 7

7

25

Using CRRs to Profit From Transmission Outages in PJM Market

• Regime 1—Little incentive to use transmission outages to cause

congestion because transmission owner had fixed price contract with retailer in DPL Zone

– No relationship between transmission outages and congestion charges to DPL Zone

• Regime 2—Incentive to use transmission outages to cause

congestion because fixed contract no longer in force and affiliate of transmission owner possesses significant quantity of CRRs into region

– Virtually all congestion charges due to transmission outages

26

Average Real-Time Hourly Values Regime 1: 04/01/98 - 07/22/99

• 1.00

4.00 9.00 14.00 19.00 24.00 29.00 34.00 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Hour $/MWh CCONG OUTCONG

Using CRRs to Profit From Transmission Outages in PJM Market

27

Average Real-Time Hourly Values Regime 2: 07/23/99 - 06/24/01

• 1.00

4.00 9.00 14.00 19.00 24.00 29.00 34.00 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Hour $/MWh CCONG OUTCONG

Using CRRs to Profit From Transmission Outages in PJM Market

28

Efficient CRR Allocations

• Efficient CRR allocation results in all load-serving

entities using CRRs as passive hedges against congestion charges not as a revenue source

– This imposes restrictions on CRR allocations depending on the local generation and transmission network holdings of load-serving entities – If all load-serving entities had no transmission or generation holdings virtually any CRR allocation would be efficient

• Would not distort behavior of any market participants

– Equity considerations can be accounted for once restrictions implied by efficiency have been taken into account

SLIDE 8

8

29

Efficient CRR Allocations

• Example of restrictions implied by efficient CRR allocation

– Two Load Serving Entities (LSE) located at end of 90 MW transmission line each has 100 MWh of load

• Assume that 90 MW of CRRs are available to be allocated

– Price outside region is $20/MWh – Retailer NLG owns no local generation – Retailer LG owns 150 MW of local generation

• 120 MW with a cost of $20/MWh
• 30 MW with a cost of $30/MWh
• Suppose that CRRs are allocated proportional to load without accounting for local

generation ownership

– Each LSE would receive 45 MW = (90 MW)/2

• Given this CRR allocation, it is profit-maximizing for LSE LG to make only 109

MW available from its low cost units and 1 MWh from it high cost units to set the local price at $50/MWh

– This implies receipt of $450 = 45*($30 - $20) in CRR revenues to offset the $10/MWh = ($30- $20) increase in the cost of generating the 110 MWh necessary to meet local demand – This yield greater profits than supplying $110 MWh from its $20/MWh units and having no congestion charge

30

Efficient CRR Allocations

• In this simple example, allocating more than 1 MW of CRRs to LG would

cause it to withhold energy from their low cost unit to profit from the congestion revenues associated with their CRR holdings

– CRR revenues = $10 with Additional Production Costs = $10 at a CRR allocation of 1 MW to LG – CRR revenues = $20 with Additional Production Costs = $10 at a CRR allocation of 2 MW to LG

• Efficient CRR allocation in this example requires less or equal to 1 MW

allocated to LG

– Remaining 89 MW or more allocated to NLG

• This is extreme example, but it illustrates how local generation ownership or

transmission ownership can lead a firm to use its CRRs to cause rather than relieve congestion

– If (local generation + CRR holdings) is greater than load obligations then it is potentially profitable to withhold output from local generation units to cause congestion and profit from CRR holdings

31

Concluding Comments

• Price differences versus costs differences determine value of transmission upgrades

in WM regime versus VI regime

– Local market power major determinant of extent of locational price differences – Cannot mitigate all or even a substantial amount of local market power – Must rely on competitive forces, which are enhanced by transmission upgrades

• Transition to WM regime implies large re-allocations of costs that may not be

justified because market is operated on network designed for former VI regime

– Further rationale for upgrades during transition period

• Limited incentives for transmission upgrades in wholesale market regime

– Suggests need for entity charged with ensuring economically reliability upgrades (that foster wholesale competition) are undertaken – Similar to CALTRANS system for California’s highway network

• Cost causation difficult to establish and cost of transmission network small

component of delivered price

– Postage stamp transmission pricing for costs that cannot be allocated on causation basis

• Congestion Revenue Rights (CRR)

– Allocate to load to protect consumers – Account for local generation holdings to protect load