Models and datasets used in Grid Investment Test analysis

The Commission wishes to provide sufficient material to interested parties so that they may replicate the process by which the analysis underlying Transpower's proposal was validated by Commission staff. A full replication of that process will require some specialised software. Unfortunately this is a necessary pre-requisite for modelling investments in the power system.

The Commission and Transpower used commercially available software:

PSS/E:  Power flow software
SDDP:   Optimal generation scheduling, taking account of hydro inflow uncertainty
Matlab:  General purpose numerical analysis

Excel was also used to verify some of the Matlab modelling.

Before submitting its Proposal, Transpower provided the Commission staff with a draft version of all economic models that it was using. Commission staff checked the draft analysis, and constructed an 'Excel' check model that duplicated most of the features of the full economic model used by Transpower. This enabled Commission staff to understand the full economic model that was going to be used by Transpower, and to verify that the calculations undertaken by that model were valid.

• Draft economic analysis and Excel check model zip [30470 KB]
  

Transpower then submitted its Proposal, and provided the Commission with its Economic Model, and input data zip [30754 KB]. Commission staff were able to replicate the Transpower analysis, but noted that there was a small error in the loss modelling (the 400kV interconnecting transformers had been modelled with an incorrect 'transformer base').

Transpower corrected this error, and it is Tranpower's corrected model which produces the final $11.1 million NPV result in favour of the Proposal.

Susbsequent to decisions by the Commission Board on a number of input assumptions to the Economic Model, Commission staff re-solved the model to arrive at the final NPV results used by the Board in making its decision.

• Modified model zip [30457 KB]

Commission staff and consultants then tested the Transpower power system analysis (PSA) by constructing PSS/E power flow cases for the 'decision points' or need dates described by Transpower in its Proposal. This was to verify that the Proposal and alternatives had been treated in an even handed manner in the PSA, and to check need dates.

• PSS/E cases  constructed to verify the PSA.

An important component of the economic case for the Proposal is the transmission losses afforded by the Proposal.  Transpower provided all files, data, and models required to calculate loss benefits. Commission staff were able to replicate this analysis, model several variants of the 220kV alternative, and validate that the loss model was solving a system that faithfully represented the Proposal. All of the files required to re-run the loss modelling can be requested from the Commission.

Next, the Commission produced ad hoc Matlab models to investigate other issues related to the Proposal, such as benefits under clause 9 of the Grid Investment Test and need dates for different investment stages.

• Ad hoc Matlab models zip [3124 KB].

The Commission constructed a reliability analysis zip [11391 KB] to verify that the need date for the Proposal was sufficiently conservative with respect to the risk of unserved energy in Auckland if the commissioning date proposed by Transpower was too late.

Return to decision page