Time to change our use of language about California’s State Water Project and the Central Valley Project.  Out with the old (supply reliability) and in with the new (risk management of water shortages).

Supply Reliability

Supply reliability relates to the ability of a water supply to meet its water demands.  Shortfalls in supply may translate into water shortages.  Water shortages have economic costs and consequences.

In engineering and supply chain management, supply reliability is measured as the percentage of the time a system operates as promised.  Water Right A yields 95% of its stated amount (80% + 20%x75%) and Water Right B yields 75% of its stated amount (50% + 50%x50%)—see table.  Are these measures an economically meaningful definition of supply reliability?

Yield of Water Rights

No.  Consider the yield of Water Right B* where shortages occur 50% of the time but the right still yields 90% during shortages.  The measured reliability of Water Right B* would be 95% (50% + 50%x90%).  Are Water Right A and Water Right B* equally valuable?  Water Right A has less frequent shortages (20%) but larger water shortages (25%) when they occur than Water Right B*, which has shortages 50% of the time but smaller water shortages (10%) when they occur.

Water Right B* has a lower expected economic loss from water shortages than Water Right A—see table below where there is a 100,000 AF water demand for the water rights.  Water Right B* has expected economic loss of shortages 40% lower than Water Right A.  For a given expected shortage, the magnitude of the shortage is more important than the probability of the shortage. 

Annual Expected Economic Loss from Water Shortages

Economic Loss: from economic analysis of Bay Delta Conservation Plan
Size of Shortage: 100,000 AF multiplied by Yield Loss
Economic Loss: Size of Shortage multiplied by Economic Loss/AF Shortage
Expected Economic Loss: Economic Loss multiplied by probability shortage

Supply reliability measures the likelihood of a water supply equal or exceeding a defined threshold.  For example, the Department of Water Resources estimates that there is a 99% probability that the yield of the State Water Project will exceed 12%.  So, with about 4.2 million AF of Table A contract amount, the amount of SWP water available will exceed 504,000 AF 99% of the time.  What should we call the reliable yield of the SWP?  What is relevant from an economics perspective is what happens during water shortages.

For a water user, at what SWP allocation do they suffer shortages, 25% allocations, 30% allocations?  What are the consequences of shortages?  How frequent and how large will the shortages be?  As shown by the above example, Water Right A would be judged more reliable than Water Right B*.  In fact, the expected economic losses from water shortages are greater for Water Right A than Water Right B*.

So, it is time to focus on the expected economic losses of water shortages.  To its credit, the economic analysis of the Bay Delta Conservation Plan does exactly that.  The proposed tunnel project does not increase the drought yield of the SWP.  Instead, it increases the SWP yield in normal and wet years.  By using under-utilized storage, the project’s water supplies can be managed to reduce the expected economic losses from water shortages.

Risk Management

The California Water Industry must stop talking about the supply reliability of the State Water Project and the Central Valley Project.  (After all, this year has demonstrated that the reliable yield of these projects is virtually nothing).  Instead, the relevant questions involve how do water systems manage the risk of these project supplies to meet their customer’s water demands, while minimizing the expected economic costs of water shortages.

Demand Reliance.  How much water demand can reasonably rely on the water source?  As explained in the first post in this series, this level of water demand is less than the average yield of a water source.  How much less depends on other circumstances.

Carryover Storage.  Storage is a critical risk management tool.  Given the variability in the SWP and CVP, carryover storage can reduce the economic cost of water shortages significantly.

Conjunctive Use of Water Supplies.  Different water supplies often have different reasons for variability in available supplies.  In Southern California, for example, Colorado River water supplies face significant less variability than water supplies from the State Water Project.  The factors driving availability are not related.  When SWP water supplies are high, for example, the Metropolitan Water District of Southern California can store surplus SWP water either directly in groundwater storage or indirectly by using less Colorado River water and storing the unused Colorado River water in Lake Mead for use in future years.

Time water managers apply tools from finance.  Investors compile a portfolio of investments to generate expected returns and manage financial risk.  Similarly, a water supplier must look at their portfolio of water assets in the same manner.  As discussed in a future post in this series, this means that water sources must be evaluated in terms of their contribution to water system performance, not in isolation.  So, for the case of the State Water Project and the Central Valley Project, their supply reliability may be zero, but their value depends on how they can be integrated with other water assets to meet customer demands while minimizing the expected economic cost of water shortages.