TrustPower defends water flow model

A computer model used to predict the impact of TrustPower's proposed Wairau hydroelectric power scheme is unreliable because the river was surveyed when flows were too high, it has been claimed.

Fish & Game lawyer Maree Baker challenged the survey during questioning of TrustPower expert witness Henry Hudson at an Environment Court hearing in Blenheim to consider whether the company should be granted consent to build the scheme.

Dr Hudson was contracted by TrustPower to design a two-dimensional flow model of the river, he told the court yesterday. Modelling predicted how a proposal would alter river features such as flow, depth and velocity with the results used to anticipate effects on habitat.

Dr Hudson conceded that river conditions during the model's development were not ideal. However, he was happy that the model was accurately predicting the distribution of flow between braids at various flows, which was its main purpose.

Modelling had become increasingly irrelevant over the six years since research was started as plenty of field data had now been collected under low flow conditions so there was less need to predict the unknown, Dr Hudson said.

An expert panel had strongly recommended the initial survey of the river be deferred from August 2003 until the following summer, suggested Ms Baker. This advice could not be acted on because he had "contractual obligations" to TrustPower, said Dr Hudson.

Delaying the survey made sense because the model would be used to predict habitat for plant, animal and fish species at the low flows which would result from the TrustPower scheme, said Ms Baker. Fish & Game expert witness John Hayes had suggested that heads of braids which controlled distribution of flow were not apparent at high flows, although Dr Hudson disputed this.

A second problem with Dr Hudson's field work being carried out during high flows was that water velocity and depth was not measured along with topographical features, Ms Baker said. This meant it was not possible to calibrate simulated flow patterns against actual river conditions on the day measurements were made. She quoted Dr Hudson's own rebuttal evidence which suggested that "despite our best intentions, we could not undertake low flow verification for weeks to months after the survey during which time the bed evolved".

Initial surveys were done in August and September 2003 when flows were so strong they caused channels to shift, Dr Hudson recalled. Riverbed changes made it difficult to calibrate the model so he returned to TrustPower asking to repeat the survey.

The company "took a few weeks to decide to double the budget and have the river resurveyed in October-December when a long-term forecast predicted flows would be well below average", said Dr Hudson. Surveys were almost complete when the river again flooded causing channels to shift significantly.

"Do you throw your hands up and walk away, start again, spend a huge amount of money to do another survey then get another lot of freshes that changes the bed," said Dr Hudson. "I don't believe I have ever seen a model that is completely error-free."

Another problem was that pulsing of river levels below TrustPower's existing branch scheme, as water was stored then released to meet peak electricity demand periods, made accurate measurement difficult. Variability in the river system was superimposed on a large amount of natural variability, Dr Hudson agreed.

A partial solution could have been for TrustPower to manipulate flow through the scheme during the survey, Dr Hudson said. However, the system had insufficient storage to do this and it would have been costly.

"More to the point, there was no point in shutting off generation to get 20 cumecs [of flow] if the background flow was 50."

Ms Baker suggested this was another reason the survey should have been done during low flows.

While Dr Hudson said he was satisfied with the accuracy of his model, Ms Baker quoted evidence from Marlborough District Council expert witness, Ian Jowett, highlighting discrepancies between predicted water surfaces and aerial photographs taken soon after the survey which could result from inaccurate distribution of flow between braids.

The Environment Court hearing results from appeals to a 2008 decision by a Marlborough District Council-appointed panel to grant the consent.

- The Marlborough Express

Newest First

Oldest First