Market impact analysis of HVDC MWh and diluted HAMI charge

Market impact analysis of
HVDC MWh and diluted HAMI
charge
December 2014
Objective
• Transpower has asked us to analyse the potential market
impact of changing the current HAMI charge to a:
– MWh charge
– Diluted HAMI charge
• Considered the potential market impact assuming
market dispatch using vSPD over post-pole 3 horizon
Aug13 to Jul14:
– System costs = estimate of supply costs (offer price x cleared
offer quantity at each injection location)
– Load market costs = aggregate costs faced by loads in the
spot market (nodal price x quantity at each offtake location)
– Generation revenue = aggregate revenue earned by
generators in the spot market (nodal price x quantity at each
injection location)
– Spot market nodal prices
Issues with status quo
Small N (12)
• HAMI concentrated
on peak injection
Disincentive on SI
generation
greater than
HAMI
• Noted by generators
• Outlined by EA
PDWP
• Observed in market
offers (~177MW)
High incremental
cost to exceed
HAMI
• Potential for larger
incumbents to face
lower incremental
cost
Impact on costs
• Potential to affect
system costs and
market prices
• Higher cost resources
are used during
peak periods
3
Trade-off with diluted HAMI and
MWh
Status quo
Diluted HAMI
and MWh
High incremental cost
of generation above
HAMI
Lower incremental
cost of generation
above HAMI
Primarily affects
trading periods with
high SI generation
Affect greater
proportion of trading
periods
Potential to increase
costs during peak
periods
Potential to increase
costs during shoulder
and off-peak periods
Simulate potential impact on market to
understand impacts
4
Modelling updates
• Several refinements have been made to model
inputs since published results  similar trend and
conclusions
• These include:
– Scaling of USI generation HVDC charges (Scaling factor =
0.1)  reduces impact of diluted HAMI and MWh charge
to USI generators
– Present value adjustments to MWh charge to account for
timings of future charges (MWh charge reduces  reduces
MWh charge from $8.4/MWh to $7/MWh
– Refined assumptions of the incremental cost faced by
generators for generation > HAMI  previously assumed
probability of two potential counterfactual scenarios 
updated to
Modelling framework and
assumptions
• Estimated withheld generation (~177MW
due to HAMI) is offered into the market
• SI generators willing to generate and
increase HVDC costs provided spot market
price greater than or equal to incremental
generation and HVDC cost
• Additional capacity is offered at the highestpriced tranch of the generator
• Sensitivity solve  additional capacity not
offered into the market at price lower than
LRMC of generation (~$80/MWh)  provide a
range on impact
Estimating the MWh charge
• HVDC MWh charge was calculated using:
– HVDC revenue requirement for 2014/15 pricing
year ($145m)
– Annual average output of SI generators1 over 5
years2 (Apr 08 to Mar 13) (15,328GWh3)
– HVDC MWh charge4 = $7/MWh
• USI generators face lower effective charge
= 0.1 x HVDC MWh charge = $0.7/MWh
1Excludes
embedded and partially embedded generators
over 5 years was chosen to reduce inter-year variability which could occur as a result of
wet and dry years
3Average energy = LSI average energy + 0.1 x USI average energy
4PV using 7% discount rate (mid-year discounting)
2Average
Generator offer adjustment –
MWh
Generator energy offer curve
$/MWh
Original offer
Updated offer
DP
MW
DP = MWh charge (PV over 5 years)
8
Incremental HVDC cost – HAMI
•
•
•
SI generator considering the marginal decision to increase output
and increase its HAMI
Marginal economic cost of the decision = incremental HVDC cost
of generating an additional unit (1MW) and forgone benefit of the
alternative (counterfactual)
Two potential counterfactuals:
– Counterfactual scenario 1 (CF1): Another generator increases output but
does not affect the HAMI  more likely with smaller N
– Counterfactual scenario 2 (CF2): Another SI generator increases output that
does affect the HAMI  more likely with larger N
Probability
of CF 1
Probability
of CF 2
CF1 incremental cost
= DCR x (1 – ki) x (1/N)
Where ki = share of HAMI
CF2 incremental cost
= DCR x (1/N)
Expected incremental HVDC cost
= Pr(CF1) x CF1 incremental cost
+ Pr(CF2) x CF2 incremental cost
Probability
of CF 2
Larger N
Smaller N
Probability
of CF 1
9
Estimated energy charge uplift to
recover expected incremental
HVDC cost
Status
quo
10
Generator offer adjustment –
diluted HAMI
Generator output duration curve
MW
Generator energy offer curve
$/MWh
HAMI
MWN
Assumed implication
of HAMI on SI energy
offers
N injections
used to
calculate
HAMI
N
n years
Original offer
Updated offer
DP
MWN
MW
N = Apr 09 to Mar 13
MWN = minimum generation used in the calculation of the
HAMI
DP = energy price adjustment to recover the full incremental
cost of generating above the HAMI (PV over 5 years)
11
Embedded generators
• Both the diluted HAMI analysis and the MWh
analysis estimate market impacts based on
potential responses from SI grid-connected
generators in response to the alternative charging
regimes, no adjustments made to embedded
generators
• Do not expect this to significantly affect the
outcome of the analysis for the following reasons:
– SI generation predominantly grid-connected (~95% of
annual SI energy production)
– Under current market rules, wind generators (embedded or
not) would not be permitted to adjust offer prices
(maximum offer price of $0.01/MWh)  (North Makarewa
and Mahinerangi ~ 2% of annual SI energy production)
– Net injection onto the grid is considered for HVDC
transmission charges  currently ~98% of HVDC charges
allocated grid-connected generators
Simulation setup to model
impact on market dispatch
• Alternate HAMI calculated as the average
(across 5 years) of N peaks where N = {100,
1000, 5000, 10000, 25000, 50000}
• Market dispatch was simulated using vSPD
over the post HVDC pole 3 period Aug 13 to
July 14 with:
– MWh charge
– alternate HAMI calculation scenarios
• Sensitivity solve  additional offered capacity
offered at the maximum of $80/MWh and
maximum offer price of that generator 
provide a range on impact
13
RESULTS
Estimated impact on system and
market costs
Increase in costs if insufficient
number of peaks in HAMI
calculation
15
Estimated impact on system and
market costs - sensitivity
Increase in costs if insufficient
number of peaks in HAMI
calculation
16
Benmore price impact
(peak and shoulder)
Highest 100 prices
If N is not large enough then can get an increase
in peak and shoulder prices  greater % of
generation at prices > peak price
Highest 101-500 prices
Increases in “shoulder” prices if N is not large
enough
• Larger N in HAMI can
result in reduction in
peak spot prices 
peak spot prices >
energy offer +
incremental cost to
compensate for HAMI
 similar effect as a
MWh-type charge
• Not increasing N
sufficiently can get an
increase peak and
shoulder prices 
incremental cost is still
too high  peak spot
prices < energy offer +
incremental cost to
compensate for HAMI
 also a greater
proportion of
generation is affected
(relative to status quo)
17
Benmore price impact
(low-priced)
Lowest 1000 prices
Increased market prices during lowerpriced periods under MWh  increases
system costs and also increases load
market costs to load and generator
revenue  similar effect if N is very large
(e.g. all trading periods)
• In lower priced periods  larger N in HAMI calculation has
the potential to increase prices relative to smaller N as
greater range of generator output affected
• Impact on increasing prices during low-priced periods likely
to be smaller than a MWh-type charge that is passedthrough as a variable cost on generator output
18
Otahuhu price impact
(peak and shoulder)
Highest 100 prices
Lowest 1000 prices
Highest 101-500 prices
• Similar effect as
observed at Benmore
19
High priced period example –
29 Jan 14 (trading period 28)
Scenario
OTA2201
($/MWh)
BEN2201
($/MWh)
Actual
1,864
1,512
MWh
1,042
HAMI (N = 25000)
1,042
•
•
•
•
System cost
impact ($k)
Load market
cost impact
($k)
845
-23
-1833
845
-23
-1833
In trading period 28 on 29 Jan 2014, actual spot price in NI and SI
exceeded $1,500/MWh
Both MWh and HAMI (N = 25000) enable dispatch of additional SI
generation since incremental cost of additional HVDC charges
(MWh or HAMI) is “economic”
Comparable system cost and load market cost reductions for MWh
and HAMI (N = 25000)
Load market cost reduction much larger than system cost since
represents reduced cost faced by all loads (there is a
corresponding reduction in generation revenue)
Low-priced period example –
05 Jan 14 (trading period 7)
Scenario
OTA2201
($/MWh)
BEN2201
($/MWh)
Actual
0.01
0.01
MWh
7.7
HAMI (N = 25000)
0.01
•
•
•
System cost
impact ($k)
Load market
cost impact
($k)
7.2
0.09
11
0.01
0
0.3
In trading period 7 on 05 Jan 2014, actual spot price in NI and SI
was $0.01/MWh
In MWh scenario adjusted offer prices sets floor price in the SI 
price in NI is greater due to transmission losses  increase in load
market costs and smaller increase in system costs relative to status
quo
HAMI (N=25000)  benefit to loads less impact in lower-priced
period (relative to MWh)  primarily market benefit to load as
system cost benefit likely to be small (flat supply curve during
these lower priced periods)
Impact of N
Low-Medium: Potential to
increase peak and shoulder
prices  less impact on lower
priced periods
N
High: Potential to reduce peak
prices but increase shoulder
and off-peak prices
Very high: Potential to reduce peak
prices but increase greater
proportion of off-peak prices  starts
looking like a MWh charge
22
Conclusion I
• As N increases:
– incremental cost of exceeding HAMI (HAMI price uplift)
reduces
– greater proportion of generator output affected
• In higher-priced periods, HAMI (large N) and MWh
have similar impact since:
– Offer price + MWh charge < high spot prices
– Offer price + HAMI price uplift < high spot prices
• MWh approach more likely to affect lower-priced
periods  increasing prices and system costs
23
Conclusion II
• MWh and diluted HAMI (large N) indicate benefits
relative to status quo
• HAMI with N~25000 over 5 years:
– Similar system cost reductions as MWh scenario
– Larger reductions in load market costs (generator revenue)
due primarily to reductions in low-priced periods, relative to
MWh
Impact
Diluted HAMI
(N = 25000 over 5 years)
MWh
System cost ($m)
-3 to -16
-4 to -17
Load market cost ($m)
-73 to -143
-52 to -128
Gen revenue ($m)
-68 to -142
-48 to -127
24
THANK YOU
QUESTIONS?
ADDITIONAL SLIDES
• Post-process calculation to
check if hydro trajectory is
feasible for scenarios
• Both MWh and diluted
HAMI have feasible hydro
storage trajectories