The focus on the cost of new power stations was brought to the fore following a talk by a well-known and well respected economist at a recent conference. The core issue here is the cost of building a new nuclear plant, which is generally considered to be “too expensive” or “unaffordable”. The term “expensive” in this context is relative in nature, and must be judged by comparison with the cost of other technologies. Specifically, in this case, other “carbon free” generation technologies.
This is where the problem arises, firstly because this comparison is never made, and secondly because there does not seem to be an accepted basis for comparison anyway. Where the cost of nuclear is discussed, the cost of the alternative is carefully avoided. Costs quoted are based on nameplate power rating and a plethora of other fictional items. Each of these is used in some way to justify a particular approach and there does not seem to be any meaningful or realistic way of comparing costs. One can surmise that the comparison is never made for the very good reason that the result runs counter to the popular belief that renewables are cheaper than other technologies.
To get to a realistic basis we should remember that we consume energy and not power, so that a fair basis for comparison would seem be the cost of plant of equal energy generating capacity. Following this reasoning gives some interesting results. For instance, wind or solar plants of equal energy generating capacity to nuclear requires three to four times the nameplate power rating capacity. To generate the same amount of energy as a 9600 MW nuclear power plant would require approximately 38 000 MW of renewable nameplate capacity. And that means three to four times the power rating based cost. Taking published costs for renewable energy plant installed under the REIPPPP gives a cost of the renewable alternative that is above the estimated R1-trillion for nuclear, ignoring the fact that this will require storage to behave anywhere like the nuclear option. This was the point which flummoxed our economist friend.
The question currently being tossed around is whether South Africa can afford nuclear. If the answer is that we cannot, then surely the other carbon free alternative, costing more, could not be afforded either. Affordability is dependent on who is going to finance the project, and there seems to be a vague belief or assumption that even if the government or Eskom could not afford to finance the nuclear option, the private sector would be able to “find the money somewhere” to continue financing renewable energy projects. After all, the sector has just invested R200-billion in renewable projects and finding six times that amount should not be difficult. This belief is bolstered by the fact that the renewable option will consist of hundreds of smaller projects, each of which will be financed individually. Maybe a reality check is needed, as the there is a limited supply of finance, whether it is channelled through government or the private sector. If the nuclear option is unaffordable then so is the renewable one, which leaves us with no other option but to drop the low carbon goal in the IRP and look to fossil fuel systems for new baseload plant.
The popular answer to this situation is that baseload is unnecessary anyway or that baseload can be simulated by a combination of variable sources and peak generation plant, often accompanied by computer modelling showing how this could be achieved. One can always achieve a reliable minimum level of generation by grossly over-installing plant, but this of course comes at a cost. Baseload means what it is at the moment and no amount of fancy modelling or hypothesis that relies on the benevolence of nature and whose design is based on some projection of future weather behaviour, can replace a controllable plant chugging away in the background generating at a steady rate. Fortunately we will never have to call on the supporters of these proposals to deliver electricity to us.
The other comparison which is often made is the unit cost of electricity to the network operator. There are two types of cost here, a regulated cost which applies to the generating utility and is based on actual annual costs, and a power purchase agreement (PPA) cost, which applies to the independent power producers. The regulated cost is based on pass-through of all reasonable costs, plus a margin, whereas the PPA cost is a price bid by the IPP.
We have seen some strikingly low costs in the REIPPP programme, but what is often not realised is that the prices quoted are indexed. An indexed price is a price which increases annually according to some index, in this case the CPI inflation rate, and can be fully indexed or partially indexed. The price that is commonly referred to in comparisons is not the average lifetime price over the PPA but the inset price. The average price can be considerably higher, as can the final price, depending on the index.
In the case of renewable energy, where the bulk of the cost is capital, using a fully indexed price allows an unrealistic low initial price to be bid, while still providing the required return over the life of the agreement. Window three of the REIPPPP called for both fully indexed and partially indexed prices, and the latter prices were significantly higher, perhaps more realistically reflecting the average cost of unit of electricity over the lifetime of the agreement. So the second question is whether we can afford to pay a continually increasing price for renewable energy, a price which is not subject to annual review by the regulator, and for a system that operates under a priority dispatch regime.
Can we afford a large portion of “carbon free” generation plant in the build program? It would appear not. If nuclear falls out of the program then renewables must go the same way and we will have to fall back on either coal, which we have lots of, or gas which we don’t. Or perhaps one of the visionaries who draw up the IRP will come up with some other viable solution.
Send your comments to: firstname.lastname@example.org