If one were to take our senior politicians at face value, one might well get the impression that a future nuclear power station construction programme is a foregone conclusion. Nothing could be further from the truth. Our own nuclear debates appear somewhat like a comet with an elliptical orbit to a pre-scientific society. We respond either with fear or veneration but it all goes away, only to return again at some unknown future date.

The more interesting question is why it does this. What is the continued attraction of nuclear? The nuclear energy sector future looks to be in a dire state. The graph below, from the International Energy Agency (page 11) shows that nuclear’s best days were 30 years ago:

There is talk of a nuclear renaissance but this is going nowhere. Most of the current new builds are in China but in 2014, China spent $9-billion on nuclear while spending $83-billion on wind and solar. French nuclear company Areva, once the standard bearer for nuclear, is now technically bankrupt. Most telling is the fact that no Generation III reactors (the ones we are apparently looking to procure) have come into service in the past 20 years due to continued delays. Outside China, the average age of the world’s nuclear reactors rises. Of the operable 437 reactors, 200 of them have been operating for more than 30 years. A third of the reactors in the US have been operating for more than 40 years and there are no replacements in sight.

A better way of understanding the attraction is to see nuclear merely as a sub-category of megaprojects. Seen in this way, Medupi and Kusile along with our World Cup stadiums are part of the same problem. Megaprojects that fail are not just a South African problem. Far from it. Bent Flyvbjerg, a management professor at the University of Oxford, proposes in a paper what he calls

The “iron law of megaprojects”. They are, he writes, “over budget, over time, over and over again”. About 90% of megaprojects experience cost overruns, the same proportion take much longer to build than budgeted and only 10% deliver on their promised benefits. Put together, success on all three criteria is one in a thousand (ie 10% x 10% x 10%). The type of megaproject is not the issue: they include tunnels, bridges, dams, highways, airports, hospitals, skyscrapers, offshore oil and gas rigs, aluminium smelters, stadiums, aerospace, particle accelerators and many others. The size and scale of cost overruns and delays on several well-known projects around the world is beyond conception. His paper has important implications. He calculates that total global megaproject spending to be between $6-trillion and $9-trillion a year, or as much as 8% of global gross domestic product. Infrastructure alone accounts for $3.4-trillion a year or approximately 4% of the total global gross domestic product and much of this will be delivered as large-scale projects, and this figure has been growing.

Why does humanity persist with all this despite the iron law?  Flyvbjerg gives various reasons.  Engineers like the technology aspects; politicians love the attention they get from these monuments to themselves; and of course the developers, bankers, lawyers, consultants, landowners, contractors, and  trade unions all get to benefit directly. Policy makers are attracted by promises of employment, the fact that megaprojects often contain a relatively high level of domestic inputs relative to imports and that once completed they can improve a nation’s productivity and competitiveness by lowering production costs. If only it was done right. But doing it right is more easily said than done. Sure, some megaprojects have been successful in at least one of their measures. If one could study what the causes of success were, one could replicate the learning elsewhere but in order to do so, there needs to be a large enough sample of successful projects to produce statistically valid answers. The problem? Success is so rare that, at present, it can only be studied as small-sample research.

What are we then left with? The answer, it seems, is a “break–fix model”. Promoters of megaprojects don’t know how to deliver successful megaprojects and so, sooner or later, they “break” when reality catches up with optimistic (or manipulated) estimates of schedule, costs, or benefits. When this reality sets in, projects are paused, restructured and refinanced in an effort to fix problems to allow delivery of some version of the initially planned project. By that time, using Lady Macbeth’s famous lines “I am in blood stepped in so far that should I wade no more, Returning were as tedious as go o’er”. So much money and effort has already been expended that it is impossible to drop projects altogether.

The Nobel prize winning economist Albert Hirschman provided some intellectual cover from this “break-fix model” in the 1960s by saying that if people knew in advance the real costs and challenges involved “they probably would never have touched it” and consequently nothing would ever be built. Hirschman suggests that humans are “tricked” into doing big projects by their own ignorance but this is a positive thing because just as we underestimate the difficulties, we also underestimate our creativity in dealing with the difficulties. We all delude ourselves with wishful thinking to make something we really want happen. This, Hirschman described as a “Hiding Hand”. It is some sort of invisible or hidden hand that hides difficulties for us and when we discover the full extent of the difficulties, these are offset by a “roughly similar” error in underestimating our ability to overcome the difficulties. One can understand the attractiveness of this type of thinking except, as Flyvbjerg points out, there is no data that support this view. Quite the opposite: The Hiding Hand is itself an example of optimism and does not capture the reality of megaproject management. Instead, behavioural theories of optimism bias, the planning fallacy, strategic misrepresentation, and principal-agent behaviour provide far better explanations to what really happens.

There are deep consequences to all of this. Underestimating costs and overestimating benefits as is typical for promoters of megaprojects means the project gets the go-ahead despite it being neither financially nor economically viable. It also squeezes out other more modest and incremental projects that would otherwise have rendered better returns had the real costs and benefits of both types of projects been known upfront. Deliberate misrepresentation of costs and benefits is an assault on democracy, good governance, transparency and accountability. The widespread practice of underestimating costs and overestimating benefits used by many megaproject promoters to convince us of the merits of their project makes it difficult to decide which projects deserve consideration. What we get are not the best projects but those that look the best on paper. In general terms, the projects that look best on paper are in fact the worst projects in terms of cost overruns, benefit shortfalls, and the risks of non-viability.

All of this gives us more than enough reason to be extremely sceptical about any claims about the proposed nuclear build programme. Nuclear falls within the broader megaproject category subject to Flyvbjerg iron law, within a sub-category of electricity infrastructure, it is the worst. This much is clear from a 2014 paper written by Benjamin K Sovacool, Alex Gilbert and Daniel Nugent entitled “An international comparative assessment of construction cost overruns for electricity infrastructure”. In it, they assess the construction costs of 401 electricity infrastructure projects built between 1936 and 2014 in 57 countries. Collectively, these projects involved about $820-billion worth of investment. They show that costs are underestimated in about 75% of projects across the entire sample. Importantly, they show that cost risks are very different depending on the type of infrastructure. One of their graphs displayed below is particularly telling:

Graph: Average cost escalation and frequency of cost overruns for electricity infrastructure projects

Nuclear’s average cost escalation is nearly 120% of costs and the cost overruns happen nearly every time. Moreover, because of the size of even the smallest viable nuclear project (a single reactor has a capacity of around 1,000MW), the absolute cost of the almost certain overruns can be catastrophic. Perhaps this is why the 2013 World Nuclear Industry Status Report states that “rating agencies consider nuclear investment risky and the abandoning of nuclear projects explicitly ‘credit positive’”. The same report mentions that of the nuclear utilities assessed, 67% between 2008 and 2013 were downgraded.

The notion that construction risk can be passed onto the vendor or entity constructing nuclear power plants is pure fantasy. As the upfront costs are so high, the capital costs have to be covered which, in turn, requires the contracting government to provide a financial guarantee. Beyond the obvious point of not having a vendor cut corners to make the budget, seeking to transfer financial risk elevates financing over the technical elements. It is somewhat similar to choosing a car based on the financing package that can come with it. Despite efforts by other countries to push financial risk onto the vendors, it has never successfully been done anywhere before.

There is no danger that South Africa will be building nuclear. Even pro nuclear commentators are highly sceptical about whether our nuclear programme will go anywhere. As we have seen, there is severe resistance to funding Eskom’s own shortfalls through increased tariffs. Adding a nuclear programme on top of this is just not viable, even if such funding were available. We know we cannot fund it directly ourselves either. We also know that the Treasury has not looked at the basic numbers yet. It is simply not credible when we say that we are in the final stages of procuring nuclear reactors when those tasked with our national budget have not looked at the numbers yet. The national government is bound by the Public Finance Management Act (PFMA) which deals in some detail with procurement. The PFMA itself rests on Section 217 of the Constitution. It is worth quoting from Section 217(1) in full: “When an organ of state in the national, provincial or local sphere of government, or any other institution identified in national legislation, contracts for goods or services, it must do so in accordance with a system which is fair, equitable, transparent, competitive and cost-effective”.

The real danger of the nuclear programme is not that we are on the verge of procuring nuclear power, it is this: it distracts us from doing those mundane things necessary to lead a recovery in our broken electricity supply system. It also holds up the development of viable alternative projects that can be built on a more distributed basis in smaller, more predictable chunks that can scale up more smoothly and with much less risk. These options exist and we should focus on them instead. DM

Photo: A worker operates in a factory for the treatment of hard radioactive waste at the Chernobyl nuclear power plant in Ukraine, 25 February 2015. Construction works for a new protective shelter are currently underway at the site of the biggest accident in the history of nuclear power generation. EPA/SERGEY DOLZHENKO.