Brian Molefe is a braver man than I am to make such a daring vow that there will be no load-shedding this winter. I have no doubt that it is a fair reflection of what Eskom plans to achieve. Very real strides have been made in bringing plant back out of maintenance and getting the first unit of Medupi fully operational in time for us to scrape through the coming winter.
Chris Herold
Eskom is packed with very smart people who appear to be of the opinion that enough to get by is enough for now. I can only applaud the efforts that have been made and I sincerely hope and pray that is exactly how it will pan out. However, this most desirable prognosis depends heavily on nothing breaking, a mild winter that doesn’t result in too heavy a demand for energy for heating and the economy continuing to stutter along at low capacity. The latter, of course, is something of a self-fulfilling prophesy. I suspect too, that we will indeed have load shedding of a form that is largely invisible to the public. Eskom has agreements with the some of our biggest industries that enable them to cut demand by up to 10% by paying them not to use electricity! Fancy that – paying people not to buy your one and only product! By dint of this extraordinary (and economically damaging) ploy it is possible to lose 10% of the power utility’s demand without Joe public experiencing any load-shedding. It is also significant double whammy to Eskom’s balance sheet. Not only do they lose 10% of their income during such events, they also lose whatever they are paying to those taking a cut in electricity supply. While by no means the only reason for the utility’s perennial cash-flow problems, it surely adds to them. Nevertheless, it does help to keep the lights on and is a pretty neat way to keep the masses happy. (Perhaps not quite so happy for the few who put two and two together and link the drag on our economy and the reduction in output to unemployment.)
Reading between the lines
What was left unsaid is illuminating. Molefe rightly mentions bringing base load generators back on line after maintenance and the commissioning of new capacity as the reasons why we may escape load shedding this winter. But solar PV and wind generation are conspicuous by their absence from the credits. Despite expenditure of R170-billion, instead of giving us the value of an entire Medupi power station (the price tags are eerily similar), the contribution of the renewables is not deemed worth mentioning. In all fairness, all of the generating capacity that the R170-billion will provide is not yet on stream. But on 19 May the minister of energy, Tina Joemat-Petterson, was quoted as saying that 1827 MW is already connected to the grid.
One sixth of the capacity of Medupi (800 MW) is rightly hailed as an outstanding step in alleviating our electricity crisis (pity it didn’t arrive five years ago as planned). But 1827 MW of renewables doesn’t even warrant a mention! Why? Did someone overlook their contribution? Of course not! How could anyone forget the R170-billion of debt that has to be paid off in the form of bloated electricity tariffs? No, all this shiny new very expensive equipment was not mentioned because there is simply too little to mention. The solar PV (half of the investment) produces nothing at all during our biggest early evening diurnal power demand peak. Wind can maybe provide 15% to 20% of its nameplate capacity. (Although on recently passing one of these large wind farms on the way to Cape Town during an early morning peak demand period I couldn’t help noticing that all but a few of the giant windmills were not turning and the one or two that were could only manage a couple of rpm. Effectively the power generation was as close to zero as makes no difference. When that occurs during peak demand periods what happens to the electricity supply? So maybe even the 15% is a bit optimistic?)
If we are really generous and assume that wind can deliver 20% assured base load, then of the 1827 MW of nominal capacity that is connected to the grid and raring to go, we get only a measly 146 MW of reliable base load, about half of the capacity of Eskom’s smallest and oldest generating set! (This calculation is based on the assumption that wind comprises about 40% of the renewables installed capacity.) Most folk would be too embarrassed to point out that despite paying a king’s ransom a measly 8% of the nameplate generating capacity is actually available to help us out of our desperate energy crisis. That’s the equivalent of R400-million per MW – about 12 times higher than the unit capital cost for Medupi power station! (Which we know is grossly over-priced, maybe by a factor of two, due to procurement problems and a comedy of errors that have set back its construction programme by no less than five years and rising.)
The bottom line is that our 90% of our much vaunted splurge on spending vast sums of money on renewables is all but useless at alleviating our severe energy crisis. So Molefe is quite right not to make mention of their paltry contribution.
Their main achievements have been to:
OK, so I have been sarcastic. All of the above are anything but achievements. They are actually sources of shame. But there are some real benefits, although it is doubtful that they come anywhere near to justifying the enormous cost:
When you cut through all the hyperbola and spin, the last two bullets are the only useful contributions.
The first of the two is tangible. However, it must be recognised that this will never be economically viable until the cost of the renewables drops below the marginal cost of the coal that is displaced. This is not likely to happen in the foreseeable future. Certainly not within the economic life of the renewables that are being invested in.
The value of the second is at best speculative. Far too many ifs and buts: If anthropogenic emissions are the primary cause of climate change. If the rate of change is as high as the GCMs (Greenhouse Gas Models) predict (they have been embarrassingly inaccurate to date). If the effects are as bad as speculated (but studies major on the downside with little attention paid to the benefits). Even more significantly for us, how do the benefits stack up against the costs? There is strong evidence that for us the cost of the cure is far worse than the bite. Worse still, no matter what we spend on reducing carbon emissions will not make any noticeable change to global carbon emissions and that we will still have to suffer and fund adaptation to whatever changes in climate occur.
The CSIR’s justification of renewables
Earlier this year CSIR researchers published a report indicating that the 1600 MW of renewables which had been commissioned as at the end of 2014 had saved us far more money than they had cost. The report is nice and meticulously detailed and looks quite plausible at face value. It makes interesting reading and I have attached the summary document.
The report claims that the saving generated by the renewables amounted to R2,42/kWh against a cost of R2,08/kWh, saving 5,3 – 4,5 = R0,8-billion. The estimated saving is broken down as follows:
Wow, that sounds great! We get it for free, with a huge profit to boot. But unfortunately the wheels come off if we look beyond what is said to what is not said. The whole premise hangs on the assumption that the recently commissioned 1600 MW of renewables was the only option that could have been implemented. But what would have happened if instead we had got 1600 MW of base generation capacity? When we ask this critically important but studiously ignored question a very different picture emerges.
Let us look more closely at each of the benefits that have been attributed to the renewables:
OCGT diesel saving
If we had 1600 MW of baseload capacity instead of renewables we could actually have saved a lot more diesel fuel because the full 1600 MW would have been available for the entire time that the OCGTs were operating. For example, during power constrained periods in 2014 the renewables made it possible to reduce OCGT operation by 1,07 TWh, giving a fuel saving of R3,33-billion. But during the same year the OCGTs clocked up 3,6 TWh, to which must be added the 1.07 TWh that the renewables provided, giving a total potential saving of 4,67 TWh, with a diesel cost of R14,5-billion. Now not all of that would have been saved if the renewables had been replaced by base load generation, since the combined capacity of the OCGT generators is bigger than 1600 MW.
Calculating the saving would require processing the entire record of the national output at a fine time increment together with details of the corresponding OCTG output. I have neither the data nor the time to mess around with this, but the following graph (taken from the CSIR report for a typical power constrained day) shows the solar PV use (yellow) and wind generation (light blue) output, both of which would have been OCGT generation(pink area) if the renewables were not available. So the total OCGT generation up for grabs is the sum of the pink, blue and yellow areas. I magnified the plot to 600% (to minimise the measuring error) and used the vertical axis scale to get the vertical distance that represents 1600 MW. This turn out to be about twice the widest width of the yellow band. It is interesting that during this typical constrained day that the sum of the yellow (PV) and blue (wind) areas seldom exceeds half of 1600 MW. This implies that for this “typical” constrained day (presumably chosen from the 365 days in the year to show the renewables in the best light), 1600 MW of base load generation that would have been fully available for the entire period could have saved about twice as much diesel fuel. If this holds true for all constrained days, then 1600 MW of base load (i.e. generated by coal, hydro or nuclear) could have saved about 1,07 x 2 = 2,14 TWh, saving R6,66-billion.
Coal saving
Our most readily available mode of base load generation is provided by coal. (We have very little more exploitable hydro capacity and nuclear takes much longer to build and is a much more expensive option. Gas turbines using imported LNG (liquefied natural gas), possibly later replaced with our own shale gas, would be even cheaper still. This is an attractive future option.
In the case of coal fired generation there is obviously no saving in coal. In fact, since this option would generate much more energy than the two renewables options, it would use more coal than the renewables displace. Hence doubling the displacement of OCGT generation would be at the expense of increasing the energy sent out by 1,07 TWh. The entire renewables energy sent out of 2,2 TWh (which included OCGT replacement and coal replacement) would also have to be replaced by coal, giving a total coal based energy output of 3,27 TWh. This would cost R1,05-billion.
Unserved energy saving
Let me say from the start that there is something seriously wrong with this estimate. The value of the unserved energy (which arose during 120 hours in 2014) has been calculated assuming a cost to users of R87/kWh. Now a figure considerably higher than the cost of the electricity is perfectly reasonable since the goods produced are worth much more than the energy input. However, the quantum is too high. During 2014 248,5 TWh of energy was produced. At R87 per kWh this would come to R21 619,5-billion. This is over five times bigger than the entire GDP of South Africa. It is hard to see how value of the value generated by Eskom’s electricity can exceed the total production of the country. Aside form anything else some users (like Sasol) generate their own power; some use liquid fuels and some have little or no dependency on electricity. It may be argued that this is the marginal cost, but this too is fallacious. Some of the most valuable uses are buried much deeper in the power stack and are protected from the vagaries of normal load shedding. Even if this is not done deliberately, when a region is cut off, everyone gets no electricity, not just the highest value users! So a R21,6-trillion loss in a year is clearly impossible. (you cannot lose what you don’t have.) Hence the value of the unserved energy claimed for the renewables has be reduced by at least a factor of five (down to R0,32-billion).
But irrespective of what the value of unserved energy is, an equivalent base load generation capacity would have at the very least matched the benefit claimed for the renewables. Moreover, at any time when the unserved power exceeds that of the power output of the renewables and the renewables fall short of 1600 MW, the base load generator would have filled in the difference. Once again I am not in a position to calculate this, but suffice it to say that whatever it is, base load generation would have saved at least as much and probably considerably more.
Overall picture
So putting it all together we have:
| Source | 1600 MW Renewables | 1600 MW Base load |
| OCGT fuel saving | 3,33 | 6,66 |
| Coal saving | 0,36 | -0,69* |
| Unserved energy saving | 0,32+ | >0,32+ |
| Total saving | 4,01 | >6,29 |
Notes: * The total coal cost of 1.05 billion must be reduced by the renewables saving of R0,36-billion to avoid double accounting)
+ The unserved energy saving has been adjusted downward to not exceed the nation GDP. Whatever the true figure is, the base load generation options would realise a significantly greater saving.
But that is not the whole story. A few years down the track the renewables will no longer be at the top margin of the plot (maybe they never should have, seeing as they are so unreliable at meeting peak demand); they will be part of the rest. But running at their best they will still only produce a paltry 2,22 TWh, out of a potential energy output from 1600 MW of 14 TWh.
(Given that the installed capacity of the renewables actually ramped up from zero to 1600 MW during 2014, perhaps we can assume that the energy output might have been doubled to 4,4 TWh is 1600 MW had been available for the entire year. But even 4,44 MW – 30% load factor – is a pale reflection of the 14 TWh potential). Running flat out new baseload capacity would provide much more energy.)
The cherry on top is the ability of the base load generator to operate at capacity 24 x 7 throughout the peak demand period when the energy production is much more valuable. The value if this has not been included in the table, but it is very big.
Finally, when comparing coal with renewables for 2014 the CSIR’s stated average cost for the renewables of R2,08/kWh says it all. Escom’s average generation cost during this time was only about R0,6/kWh, and this included the cost of providing peaking power (think expensive pumped storage schemes, OCGTs and excess capacity kept in reserve to take up the peaks), building and running the grid and administering everything. The R2,08 for renewables does not provide any of this. In fact most of it disappears entirely during peak conditions! So if the renewables had to provide for base load during the peaks plus the additional costs of providing for the peaks themselves, then the cost would be very much greater than R2,08! In fact the true cost of renewables cannot fall lower than the cost of the base load generation required to replace it, plus the cost of the renewables, less the marginal cost of coal. Not unless renewables can provide energy storage to fully bridge the peak demand periods. Aside fo r the capital cost of the storage, this includes the need to drastically reduce the power sent out (i.e. sold) to pack enough energy into the storage to replace the very large energy losses and restore the base load during the peaks. Unfortunately that is still a very distant dream.
Yes, the unit cost has fallen, but it would have to fall very much more before it comes near to being competitive. Sorry to burst bubbles, but in all conscience I have to do so. We cannot afford to play unaffordable games when the economy of our entire nation is at stake.
Contact Chris Herold, Umfula Wempilo Consulting, Tel 011 462-1517, chris@herold7.co.za
