Climate change will change the long-run cost of any given amount of water supply.

The Western Australian economic regulator has adopted the perturbation or ‘Turvey’ approach to the calculation of long-run marginal cost (Economic Regulatory Authority 2005). The approach accounts for the nature of water-supply infrastructure, including lumpiness in augmentation and jointness in integration, as well as the inevitability of supply expansion. Turvey (1976) suggested that long-run marginal cost for water is a dynamic cost which relates to the effect of demand growth on the rate at which the water-supply system must expand. The perturbation method works out the capital-expenditure program required to meet forecast growth in demand and compares it to a second capital-expenditure program associated with a smaller (or larger) forecast pattern of demand. Long-run marginal cost is calculated as the difference in the net present value of costs, divided by the difference in net present value of quantity demanded for the two scenarios. The Turvey approach was used to assess the potential impact of climate assumptions on the calculated LRMC.

For this exercise, the baseline forecast demand was based on the Water Corporation’s per-capita demand assumption and assumed to grow with population growth. Growth assumptions in the medium term were based on demographic forecasts (Western Australian Planning Commission 2005). Beyond 2031, growth was assumed to continue at the same growth rate as in 2031. The calculations start in 2011, when the first feasible capital expenditure can occur, which is assumed be the desalination plant at a cost of $955 billion (Western Australian Government 2007). Subsequent augmentation is based on the source developments^[7] costed in Water Corporation (2005), with costs measured in 2011 dollars for consistency. A 100-year capital-expenditure program was examined, with desalination as the augmentation source after available conventional sources have been exhausted. The operating cost of desalination is assumed to be $1.04 per kL, which is significantly higher than the (Water Corporation 2005) estimate of 40–44 cents per kL, not just because of sharp increases in nominal energy prices over the time period, but also because of the fact that the cost quote for the Kwinana plant does not allow for the cost of renewable energy.^[8] A discount rate of 6 per cent was assumed.

Results are presented in Table 4. Compared to a 1975–2006 climate scenario, which provides an LRMC of $0.78 per kL, the LRMC associated with the Water Corporation’s 2001–06 climate scenario is $2.19 per kL.

These costs can be compared with the prices set recently by the economic regulator in Western Australia (Economic Regulatory Authority 2008). These prices are based on an earlier price determination (Economic Regulatory Authority 2005) that had been adjusted for inflation. The previous enquiry based the LRMC calculation on assumptions that are now out of date (including the assumption that the South West Yarragadee would go ahead^[9] and yield more than 45GL, and the assumption that climate would be somewhere between post-1975 and post-1997 historical means^[10]).

Projected nominal urban-water prices for 2011 are shown in Figure 1. Because of the inclining block tariff structure it is necessary to overlay the graph with a depiction of the pattern of household consumption. On the right-hand axis is the cumulative distribution of household consumption, indicating the proportion of households whose consumption is less than, or equal to, the annual volume shown on the horizontal axis. Around 93 per cent of households consume 550kL or less, and are therefore subject to prices of $1.03 per kL or less at the margin. This is less than half the long-run cost associated with a climate-change assumption adopted by the Water Corporation to justify its capital expenditure on another desalination plant.

 

The establishment of correct price incentives will therefore require a doubling of prices. Even if consumer demand is relatively price inelastic, a significant impact on per-capita consumption is likely. For example, it is well established that demand is more inelastic for indoor uses, with indoor (or winter) elasticity generally less than -0.05, and outdoor demand elasticity can be higher, such that measures of aggregate household demand elasticity have been reported to be around -.3 to -0.4 (NERA 2001; Arbues et al. 2003; Dalhuisen et al. 2003). About half of water is used for indoor use in Perth. Using an indoor elasticity of -0.05 and an outdoor elasticity of -0.2, the calculated impact of a doubling of prices is 15 per cent or 23kL per capita.^[11]