Archives

Categories

Investing in infrastructure with the possibility of severe future climate change

Suppose government is thinking of building a desalination plant to deal with future possible severe climate change.  Despite the current sound water supply situation over the next 10 years or so this is a reasonable supposition. Suppose further that the expected future discounted benefits of such a project are B(t) and the expected construction and operational costs are C(t) when the expectation is taken at future time t.  Should the government undertake the project now i.e. at t=0?  Using expected values standard cost benefit analysis says do so if and only if B(0) > C(0) so expected discounted benefits exceed expected discounted costs. Real option theory says that if the investment is irreversible – plausibly one cannot turn a future uneconomic desalination plant into an amusement park or a university so the investment is irreversible – and if the government gains  a better grip on the character of the uncertainty as time progresses then the  rule for developing now is tightened to become B(0) > C(0) + QOV where QOV is the positive quasi-option value of the project.  QOV is a premium that is added to costs to reflect the fact that if the project is a dud the government is stuck with it because of investment irreversibility but if the government defers the project now it might, as it learns about the future uncertainty, revise its initial decision and eventually proceed with the project at a future time.  The asymmetry of the possibilities here means the government has a propensity to be cautious about the project and to defer it.

I have been thinking about this line on inquiry for a couple of years but with an uneasy awareness that this did not closely match the recent experiences with desalination around Australia. Australia now has desalination plants in all its capital cities but in many cases benefits did not seem to match costs let alone to cover costs plus a QOV. There are plenty of smart state government bureaucrats out there so I have always wondered what was happening.

One might alternatively suppose that state governments are very risk averse with respect to extreme drought events – this might reflect real angst on the part of the electorate and might not be purely a cynical political motive.  Governments might expect to enjoy very short lifespans were a  city like Melbourne or Sydney to face really severe water shortages.   Of course they could rely on backup groundwater supplies (also an expensive backup option) or rural-urban water trades but the latter, in any event, is rainfall-dependent and may not represent an uncorrelated supply option relative to conventional streamflow driven supplies.  Real fear of severe drought might be incorporated into a real options framework by  representing water benefits as a utility of water supplies which was highly concave at low water supply levels representing high risk aversion at these levels.

Then the development rule might be something like (I have not formally derived the correct rule) C(0) – B(0) < I or  B(0) > C(0) – I where I is the cost of the optimal insurance one obtains by using a rainfall-independent technology.  Now expected benefits would not need to cover expected costs – they could even fall short of costs incurred. In addition, contrary to the real options approach, the decision to initiate the project would tend to be brought forward rather than to be deferred. The question then is whether or not desalination can provide this type of insurance at a reasonable cost.   Using figures cited in the recent Productivity Commission report on urban water, costs for Melbourne and Perth of using desalination were around $200 annually per household of 4.  My view (contrary to much conventional wisdom) is that this is a reasonable insurance premium for being protected from the effects of drought.

Two comments.  With the current desalination plants in place the insurance benefits from further use of this technology will be lower and it will be sensible to choose options that diversify against other sources of risk rather than rainfall/streamflow uncertainty.  Energy price uncertainty is something desalination is particularly exposed to because of heavy carbon input components.  

Almost inevitably comment on these claims will talk about the likelihodd that  the Wonthagii, Brisbane and other desalination plants will continue to remaining unused because of currently full dams. But there will be future droughts and water demands will probably grow as urban populations continue to grow. From an insurance perspective to say that these options are bad now because the Millenium Drought has been followed by a couple of years of record rains is like saying that the fire insurance a citizen paid for last year was a waste of money because their house didn’t burn down during the last year.  That reasoning is, of course, not an open-ended endorsement of desalination at all – it all hinges on the costs and benefits of the implied insurance that is provided. If the costs of fire insurance or of desalination are too high then it is preferable to live wityh the risk.

All the signs are that the world will heat up dramatically over coming decades.  The situation really is perilous and I am becoming increasingly pessimistic.  My view is that seemingly somewhat expensive policies to insure our communities against the prospects of severe water shortages are worth thinking about carefully. (1503)

3 comments to Investing in infrastructure with the possibility of severe future climate change

  • JB Cairns

    desal water uses too much energy and costs far too much, recycled water doesn’t as Singapore attests.

  • conrad

    I agree with the cost argument as well. To me it’s a good idea to insure against drought (even without climate change in Aus), but it was crazy to have done it using such an expensive method, which as far as I can tell, was simply done to pander to people’s ignorance.

  • Nottrampis

    Harry,

    Have you caught up with this paper.

    excuse the blatant plug but this way you know who I am and the paper is in the link.

    Rather long but damned interesting

Leave a Reply

  

  

  

You can use these HTML tags

<a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>