Electricity transmission pricing: getting the prices good enough

1
Electricity transmission pricing getting the
prices good enough?

• Richard Green
• Institute for Energy Research and Policy

2
Transmission pricing

• Geographical differentiation in the wholesale
  market
• Prices for connecting to and using the
  transmission network

3
Six objectives

• Promote the efficient day-to-day operation of the
  bulk power market
• Signal locational advantages for investment in
  generation and demand
• Signal the need for investment in the
  transmission system

4
Six objectives

• Compensate the owners of existing transmission
  assets
• Be simple and transparent
• Be politically implementable
• Green (Utilities Policy, 1997)

5
Three approaches

• Ignore transmission issues
• Ignore transmission issues, then bribe market
  participants to sort things out
• Integrate transmission issues into your market
  design(s)

6
Major power flows
Source UCTE
7
Major power flows and congestion
Congested 26-75 76-99 100
Source UCTE
8
If costs differ between areas
P
P
GW
GW
Xpts
9
If costs differ between areas
P
P
GW
GW
Xpts
Mpts
10
If costs differ between areas
and the lines are too thin
P
P
GW
GW
Xpts
Mpts
11
If costs differ between areas
and the lines are too thin
P
P
T
GW
GW
Xpts
Mpts
12
If costs differ between areas
and the lines are too thin
you could still ignore the problem
P
P
GW
GW
Xpts
Mpts
but someone will want money to sort it out!
13
Zones in the NEM

• NEM runs nodal model and dispatches according to
  nodal conditions (prices)
• Generators / loads grouped into regions
• All generators in a region receive the regional
  reference price
• Marginal cost at a reference node
• No compensation for constrained running

14
From a line to a network

• Electricity will flow along every path between
  two nodes
• It cannot be steered
• If one line fails, the flows instantly change
• Overloading any line can be catastrophic

15
(for example)
16
The impact of loop flows
A
B
C
17
The impact of loop flows
A
B
C
18
Nodal prices

• Set price of power equal to marginal cost at each
  point (node) on the network
• Marginal cost of generation (if variable)
• MC of bringing in power from elsewhere
• Centralised market uses the nodal prices
• Bilateral trades which move power pay the
  difference in nodal prices

19
Nodal trading

• Price at A 20, Price at B 30
• I sell at A, I receive 20
• I sell at B, I receive 30
• I generate at A and sell at B, I receive the
  agreed bilateral price and pay (30 20)
• I generate at B and sell at A, I receive the
  agreed bilateral price and pay (20 30)

B
A
20
The impact of loop flows
and constraints
A
B
6 MW at C needs 3 MW from A and 3 MW from B
C
21
Prices constraint AB

• Price at C (Pa Pb)/2
• 1 MW extra capacity allows 1.5 MW from A to
  replace 1.5 MW from B
• Shadow cost of constraint 1.5 (Pb Pa)
• If Pa 10, Pb 30
• Pc 20, shadow cost 30
• Pc Pa 1/3 shadow cost Pb 1/3 shadow
  cost

22
The impact of loop flows
and constraints
3 MW at C needs 3 MW from A
and 6 MW from B
23
Prices constraint AC

• Price at C 2Pb Pa
• 1 MW extra capacity allows 3 MW from A to replace
  3 MW from B
• Shadow cost of constraint 3 (Pb Pa)
• If Pa 10, Pb 30
• Pc 50, shadow cost 60
• Pc Pa 2/3 Shadow cost
• Pb 1/3 Shadow cost

24
The impact of loop flows
and constraints
3 MW at C needs 6 MW from A
and 3 MW from B
25
Prices constraint CB

• Price at C 2 Pa Pb
• 1 MW extra capacity allows 3 MW from A to replace
  3 MW from B
• Shadow cost of constraint 3 (Pb Pa)
• If Pa 10, Pb 30
• Pc 10, shadow cost 60
• Pc Pa 1/3 shadow cost Pb 2/3 shadow
  cost

26
Summary
27
Implications

• Nodal prices can vary significantly
• Over time
• Over space
• The first creates a need for hedging
• The second makes it harder
• The prices may be counter-intuitive

28
How to hedge

• Transmission Congestion Contract
• Spatial contract for differences
• Pays the holder the difference in nodal prices
  between two specified points (from A to B)
• Price at B Price at A
• Perfect hedge if you generate that amount of
  power at A and sell it at B
• Remember the real-time charge is (PB PA)

29
Whod sell that hedge?

• The spot market charges raise a surplus
• Who gets it?
• If the Transmission Congestion Contracts
  allocation is feasible, Hogan (1992) shows spot
  market surplus TCC payments
• Organisation receiving the spot surplus can issue
  TCCs and find itself hedged!

30
Inferior ways of hedging

• Financial Transmission Rights (options)
• Only pay out when value is positive
• Payments may exceed spot revenues
• Physical Transmission Rights
• Limited by system capacity
• If line limit on AB is 100, can only issue 100
• With TCCs, 100 BA allows an extra 100 AB
• Smeared share of congestion revenues

31
What if you get it wrong?

• Operational difficulties
• PJMs first market
• Economic operating mistakes
• Investment mistakes
• At present, we dont know much about these

32
How much does it cost to get it wrong?

• Compare demand and operating patterns with
  different pricing rules
• Model applied to England and Wales, 1996 data
• Numbers are country- and time-specific
• Approach is general

33
The model

• NGC system in 1996/97
• Thirteen zones (two pairs of NGCs zones are
  combined, one zone split into two)
• Iso-elastic demand in every zone
• Generation in most

/MWh
Gas, Coal, Nuclear Oil
GW
34
Transmission system model
North
A DC load flow model with losses (proportional to
the square of flows) and constraints on the total
flows across NGCs system boundaries
South-West
35
Three pricing rules

• One price for generation and for demand in each
  zone (optimal)
• One price at each node for generation, but a
  common national price for demand
• One national price for generation and one
  national price for demand (actual system)
• Constraints are managed via payments for
  constrained-on and constrained-off running

36
What is welfare?

• NGCs operating surplus
• Kept the same under each of the rules
• Generators operating surpluses
• Energy revenues less variable fuel costs
• Gas contracts assumed not to be variable
• Consumer surplus
• Area under their demand curve and above the price
  they actually pay

37
Prices winter peak
38
Prices summer trough
39
Basic results
40
Intuition for the results

• Adjustments to generation for constraints have to
  happen, whatever the pricing rule
• Here, these are in the same direction as the
  economic response to marginal losses
• Cost differences at stations partially offset
  marginal transmission losses

41
Market power

• Sometimes a problem in this market
• General incentive to raise prices
• Particular incentive to raise prices in
  import-constrained area
• Uniform pricing gives incentive to reduce prices
  in export-constrained area
• Model two strategic generators plus fringe
• Both firms change slope of bids (by region)

42
Generators capacities
North
South-West
43
Prices winter peak
44
Prices summer trough
45
Prices zone 12
46
Prices zone 1
47
Market power
48
Conclusions of this study

• Optimal pricing would create winners (northern
  consumers, southern generators) and losers
  (northern generators, southern consumers)
• It would be less vulnerable to market power
• Welfare gains of 1 of turnover are quite large
  as Harberger triangles go!

49
Other transmission charges

• Connection assets local costs
• Capacity-based use of system
• Affect investment decision, not operating choices
• Output-based use of system
• Affect operating choices and might be used to
  offset consistent errors in the market rules
• Contracts for constrained running

50
Interactions between charges

• Investing generators should consider both spot
  market and transmission charges
• With the right spot signals, transmission charges
  should be uniform
• Differentiated transmission charges needed if
  spot prices send inadequate signals
• Using both would over-signal, reducing
  transmission costs, but raising generators

51
Conclusion

• For major changes, transmission charging creates
  well-informed winners and losers
• Gains typically small relative to transfers
• With good operators, the system is resilient to
  poor rules
• Better rules will create gains worth having

52
(No Transcript)