Conclusion

Water storages play a vital role in the supply of water for irrigation farms. Storages serve to smooth variation in the supply of water and equalise the marginal value of water over time. The management of these storages is an important but difficult task. Determining what proportion of available water to store for the future, and how much to consume now, is a complex problem given the presence of substantial uncertainty over future inflows and water demands.

In Australia, major irrigation water storages are centrally managed via the announced allocation system, where each season a water manager determines the amount of water available for use now (water allocations) given prevailing storage levels. Under certain conditions a centralised approach could achieve an efficient allocation of water resources; specifically, if the water manager had perfect information on the water demand preferences of irrigators and there existed an efficient (costless) market in water allocations.

In practice, the water manager is unlikely to have perfect information on the water preferences of irrigators. There is likely to be asymmetric information; irrigators are likely to know more about their water demands than the water manager. Also, there are likely to be significant transaction costs in water trade. A centralised announced allocation approach relies heavily on trade in water allocations to allocate water between irrigators with varying reliability preferences. Given these practical difficulties, a decentralised approach, where irrigators are enabled to make there own storage decisions, may be preferable.

To demonstrate the potential costs of inefficient storage management, an economic model of the water storage problem facing a representative irrigation system was developed. This model was applied to a case study region, the Murrumbidgee. Model parameter values were set with reference to historical data and estimates from econometric literature. Using the model, a suboptimal aggressive release rule was compared with a theoretically optimal release rule.

The estimated optimal release rule generated a small reduction in mean water use in turn for a substantial increase in mean storage reserves. The model demonstrated the ability of the optimal policy to lead to an increase in mean irrigator incomes and a substantial reduction in variability of incomes. The model estimated an increase in the mean economic value of water of 11.8 per cent and a reduction in variability of more than 63 per cent. The model also demonstrated that the gains from optimal storage management, both in terms of the mean and variability of incomes, increase substantially as water availability reduces.

In this report two decentralised approaches to storage management were considered in detail: carryover rights and capacity sharing. Carryover rights have the potential to overcome some of the problems of centralised storage management. However, carryover rights are an incomplete solution, since they do not define explicit property rights to storage capacity or to losses associated with storage. As a result, carryover rights generate external effects, where individual irrigator carryover decisions affect other irrigators in the system. In an attempt to minimise these external effects, significant restrictions are often placed on carryover rights which further weaken their effectiveness.

Capacity sharing is a property rights system proposed by Dudley (Dudley and Musgrave 1988), which involves redefining water entitlements into separate storage capacity rights and water/inflow rights. Unlike carryover rights, capacity sharing ensures that storage space is efficiently rationed and that losses are internalised. Capacity sharing has a number of other potential benefits relative to systems of carryover rights. Capacity sharing replaces the traditional announced allocation system and in doing so removes a layer of regulatory uncertainty from existing water entitlements. Capacity sharing involves redefining water rights at the source which creates a number of potential efficiency improvements, including the potential to internalise water delivery losses.

One complication with capacity sharing is the occurrence of internal spills – where individual water accounts reach capacity and forfeit their inflows to other water users. However, the allocation efficiency implications of internal spills are negligible and in practice internal spills are likely to occur infrequently. Another important consideration in the transition to capacity sharing will be to minimise any actual or perceived distributional effects, by ensuring the newly defined capacity share water entitlements adequately preserve all existing irrigator water entitlements.

Capacity sharing is typically considered in the context of relatively simple water supply systems, where all water is sourced from a single storage. While there may exist some concern about the suitability of capacity sharing in more complex systems, it is not obvious that the concept could not be sufficiently generalised. The ability for the capacity sharing framework to be applied to a range of more complex water supply systems remains a subject for potential future research.