It is difficult to tell a simple story of the water-balance situation in Western Australia because of inconsistency between the governing agencies. On the demand side, the reality is that water is too cheap and demand is too high; and even considering the fact that consumers are on two-day per week sprinkler restrictions, unrestricted per-capita urban demand would probably be lower than current demand if price signals were corrected.^[2] On the supply side, one of the major sources of water for Perth, the Gnangara Mound,^[3] is in a serious state of degradation, and the Department of Water has been found by the Environmental Protection Authority to be in breach of Ministerial conditions regarding groundwater levels that were put in place to protect water-dependent ecosystems (Environmental Protection Authority of Western Australia 2007a and 2007b). The Department of Water (Department of Water 2007) has adopted the position that groundwater decline on the Mound is largely due to climate change (and therefore not their fault), while at the same time the amount of water allocated from the Mound in recent years has been allowed to increase to compensate for the climate-change impacts on surface-water reserves. That is, one of the key responses to the recent water shortages in Perth has been to mine the Gnangara Mound, rather than to reduce allocations in response to a drying climate. The Gnangara story is more complex again, because the Water Corporation is the highest-value user on the Mound and only represents 48 per cent of consumptive uses (Marsden Jacob 2006). With stronger governing institutions the amount of water allocated to the Water Corporation in recent years may well have increased, but in lieu of other consumptive uses rather than environmental uses.

The focus of this paper is on the decision-making of the water utility; and therefore, rather than focus on the optimal resolution to the above issues from a social point of view, the approach taken is to use the planning context faced by the Water Corporation.^[4]

The mean water-balance situation, given current institutional arrangements, is shown in Table 1 for the four climate scenarios discussed in the previous section as well as the complete (post-1911) historical record. The data in this table reflect average system yield and provide a simple exposition of the impact of climate-change assumptions on system yield and the planning deficit; hence the pressure for early augmentation. Note that the difference in system yield between the best- and worst-case climate assumptions in Table 1 is 60GL, which is a large amount of water by urban planning standards — for example, the capacity of the desalination plant at Kwinana is 45GL.

The investment in the first desalination plant, at Kwinana, was prompted by the acknowledgement that the complete historical sequence was no longer relevant for planning purposes. Analysis by the IOCI had shown a dramatic change in weather in the post-1975 sequence and the difference in yield between historic and post-1975 climate has a dramatic impact on yield.^[5] In the absence of this first desalination plant, urban supply was insufficient to meet projected 2007 demand for all scenarios except the post-1975 climate. The additional 45GL provided by the investment in desalination at Kwinana has provided sufficient (average) capacity to meet 2007 demand levels for all climate scenarios. However, by 2010 demand growth would be sufficient to cause a significant deficit in capacity under the six-year climate scenario. This climate scenario is necessary to justify the current decision to construct a second desalination plant to augment water supply in 2011. If the IOCI predictions are taken to be the ‘best science’ relating to the matter, there is no justification for augmentation in 2011.^[6]