Classificatio
DRAFT – FOR REVIEW
NZAS electricity allocation factor
Summary
Purpose
We have been asked to do a rapid review of Concept Consulting’s approach to estimating an electricity
allocation factor for NZAS supply contracts and consider whether any of the Energy Link assumptions and
input data would have been more appropriate to use than Concept’s.
Questions
1) Are the Concept Consulting analysis and assumptions reasonable?
2) Is there anything in the Energy Link material that is both substantially better and would make a
material difference to the outcome?
Short answers
1) The Concept Consulting analysis and assumptions are reasonable and compelling.
2) In one regard, Energy Link’s assessment may be substantially better and may make a material
difference:
a. we are inclined to view favourably their estimates of the rate of passthrough of
emissions prices to electricity prices, at least for short-term estimates, because in
principle the type of model used should be better at gauging these effects
b. Concept’s estimates are, in this regard, only one set of assumption-led numbers, so
ought not be considered the sole arbiter of an appropriate estimate of the rate of
passthrough of emissions prices to electricity prices, i.e. the Energy Link numbers
provide a useful sensitivity test
c.
we caution that we are not suggesting that Energy Link’s estimate of the overall
electricity allocation factor is better:
i. the rate of passthrough of emissions prices to electricity prices is only one part
of the calculation of an appropriate electricity allocation factor
ii. Energy Link’s overall assessment has limitations and aspects which we find
curious departures from history and expectations and those departures are
not well explained.
Below we discuss the reasoning that led to these answers.
1
W W W . S E N S E . P A R T N E R S
W E L L I N G T O N , A U C K L A N D
Classificatio
Conceptual approaches
The right questions is: what happens if Tiwai closes?
Concept takes the view that the relevant benchmark, for deciding an electricity allocation factor for the
NZAS contracts, is the price that Contact and Meridian would expect to get for their South Island generation
in the wholesale market if the smelter shuts down.
We wholeheartedly agree that view. The key questions are:
•
what was in the minds of the companies when they agreed the contract price
•
what cost of carbon was embedded in the contract price.
Clearly, Meridian and Contact would have been thinking about the money they would lose if they don’t
strike a deal. So, analysis must focus on South Island prices if the smelter shuts down. From there, it is a
question of estimating those prices with and without emissions pricing.
Given our strongly held position on this point, the rest of our comments focus solely on estimates of the
effects of emissions prices on wholesale market prices when Tiwai leaves.1
Mercury energy’s contract
We have not focused on the assessment of the Mercury contract as it is not a large part of the EAF. We see
no problem with Concept’s approach to identifying the cost of carbon that is due to Mercury firming its
prospective South Island wind generation to supply NZAS. A point of judgment is the assumed 50/50
probability of the project going ahead in case of the NZAS exit. A lower probability of it going ahead in case
of exit would raise the EAF estimate.
Reconciling the different numbers
There is a substantial difference between Concept’s and Energy Link’s respect estimates of the carbon costs
in the Meridian and Contact contracts: 0.089 and 0.21 respectively.
We cannot provide a blow-by-blow account of the details that create these differences. We can
•
observe high-level differences in the two assessments
•
gauge how much these differences matter for the end result
•
provide an opinion on respective strengths and weaknesses.
Our approach here is to focus on the issues that stand out – that are material. It is not to judge whether one
particular set of input assumptions is more apt than another.2
Summary of key differences
1 This judgement is about economic and commercial logic. We are making no judgement about legal or procedural
matters.
2 For example, long term growth in electricity demand is 1.2 percentage points higher in Concept’s analysis than in Energy
Link’s analysis. This is a massive difference. However, we cannot easily discern the effects of that difference, in the time
available. Though we can say that we find Concept’s numbers more plausible in light of the trend towards electrification
of energy supply in New Zealand.
2
Classificatio
In summary, there are two important high-level3 differences driving the two sets of estimates.
•
the rate of passthrough of emissions prices to wholesale electricity prices; Concept has a lower
passthrough rate, which is associated with a lower electricity allocation factor
•
the impact of Tiwai closure on South Island electricity prices; Energy Link has smaller impacts,
which is associated with a higher electricity allocation factor.
The table below summarises our first order4 estimates of the effects of these factors on estimates of
electricity allocation factors. We tested the sensitivity of the respective estimates by swapping the values
used by the modellers e.g. swapping Energy Link’s passthrough rates for Concept’s, leaving other aspects
unchanged.
TABLE 1 MATERIALITY OF KEY DIFFERENCES IN ESTIMATES OF ALLOCATION FACTORS
Concept
Energy Link
Report estimates
0.09
0.21
Sensitivity tests:
Pass-through rates
0.11
0.13
Effect of Tiwai exit on South Island prices
0.10
0.14
Different views on rate of passthrough of emissions prices
Both sets of estimates are affected by underlying assumptions about the rate at which carbon prices are
passed through into wholesale prices.
Concept’s electricity prices are less sensitive to carbon prices than Energy Link’s, which will make their
estimates of the electricity allocation factor smaller than Energy Link’s. This appears to be true regardless of
what one believes about carbon price levels or future changes in the market like Tiwai closing or higher
rates of electrification.
The size of this difference is material – Energy Link’s electricity allocation factor for 2024 is around 30%
higher than Concept’s (0.53 versus 0.41).
This difference is not because Energy Link thinks more fossil-fuelled generation will be needed. On the
contrary, Concept’s modelling tends to have more fossil-fuelled generation.
We surmise that the difference lies in assumptions about how emissions prices affect prices offered by
electricity generators. For example, the extent to which hydro generators adjust their offer prices to reflect
(match) the higher cost of fossil fuelled generators with emissions pricing.
In principle, Energy Link’s estimates of these effects should be more accurate than Concept’s, at least in the
short term. This is because Energy Link’s modelling offers a more detailed account of generator behaviour.
Note that this observation is relevant for estimating the component of prices that is due to emissions prices.
It does not speak to the robustness of forecast price levels (discussed further below), which comprise much
more than just emissions prices.
3 These are results of the modelling that summarise underlying propensities of the model. They are not input
assumptions.
4 Meaning these are checks of the numbers, for materiality, without taking account of feedback effects which, though
they may be material, cannot be done without re-running respective models.
3
