Submitted on April 21, 2020

A thought on everyone’s mind these days, or at least those involved in electricity supply, is what is Covid-19 doing to electricity demand?  Well, the answer to this question is not simple, and we are working on it, but the answers will come highly qualified by the uncertainty surrounding the disease in New Zealand, the economy in New Zealand, and the global economy. 

By Greg Sise, 21st April 2020

In this post I take an informal look at the first part of the question – how will the disease play out in New Zealand?   I say informal because I am neither epidemiologist nor virologist, but like most of us I have learned just enough to be ‘dangerous’ with my newly acquired knowledge.  Nevertheless, I take a look at some of the data to see what this tells us about how electricity demand might return to some sort of normal, given that it is currently down on last year by about 15% on weekdays and 10% in the weekend.

Of course, we now know that our (hopefully) last ever day at Alert Level 4 will be Monday 27th April, and that it is possible we could be moving to Alert Level 2 mid-May.

We’ve probably all seen charts that look something this, a concept that originally came from the Financial Times, which shows various quantities against the number of days since a particular threshold was reached, in this case the total number of cases per million of population.  New Zealand is shown as the red trace and we started out higher than many developed nations, but since commencing lockdown we have brought the number back to just under 300 per million, which is comparable to Australia* (261) and South Korea (206).  The data in the chart comes from the Ministry of Health for New Zealand and the European Centre for Disease Control for all other countries.  The vertical scale is logarithmic. 

For the first 19 days we were on a similar trajectory to Sweden, which has continued on a higher trajectory and now sits at around 1,300 cases per million population:  if we had continued on a similar trajectory then we’d have 6,500 cases by now, with somewhere close to 100 deaths, and still climbing steadily.

The lowest case rate on the chart is Taiwan, which is holding steady at an average of 3 cases per day over the last two weeks, and 16.6 cases per million of population.  From what I can gather, Taiwan is the best model for where New Zealand is heading, with daily cases being kept in the low single figures for the duration.  This is called an “elimination” strategy and Taiwan is possibly the only country that has achieved this, so far.  It is not total eradication of the disease, but the daily number of cases is so low, that it can be kept completely contained through the measures that we have developed in New Zealand, including quarantine for overseas arrivals, extensive testing and highly effective and fast contact tracing.

You don’t have to look too far into the epidemiology of viruses to discover that there are potentially huge complexities around how data is collected, processed and reported, and how modelling is undertaken.  It appears that many countries are undercounting the number of cases and deaths due to covid-19, simply because they don’t test to the same level, or because they don’t verify all deaths as either being covid-19 or otherwise.  So we are probably doing better in relative terms, than it appears in the chart above.

This next chart comes from a web site called Our World in Data which sources most of its data from the European Centre for Disease Control, and it shows the number of tests per confirmed case of covid-19.  Notice that New Zealand is in the top three, and just behind Taiwan, with Vietnam the leader (though it only has 106 cases).

Our relatively high rate of testing also raises the possibility that our total number of cases per million is higher than many other nations because we have a much better testing regime, and therefore we miss few cases.

Which brings us to the key question for electricity demand in the short to medium term:  are we on track to be the “next Taiwan”?

The chart below shows New Zealand’s daily confirmed plus probable cases, along with two exponential curves fitted to the data:  one for exponential growth, on the left, and one for exponential decay on the right.  Until we moved to Alert Level 4, lock-down, daily cases doubled every 1.9 days, although this included cases arriving from overseas, the rate of which fell after 26th March.

Diseases that spread exponentially also go away exponentially, either as herd immunity develops or once measures are implemented to slow the spread: the right-hand decay curve has a half-life of 3.7 days.  At this rate, by 27th April we will be down to 1.4 cases per day on average.  If the curve is adjusted to fit the data a little better over the last week, the half-life increases to 4.2 days, but we still get to 2.3 cases per day by the 27th.

What starts to become very significant at these low daily case numbers is the number of cases that arrive from overseas, or arise while arrivals are in quarantine, which has averaged 1.4 per day since the number fell below 5 per day on 12th April, and in the last few days averaged 2 per day.  So, while the total number of new cases yesterday was 9, there were 2 cases due to international arrivals, which reduces our internal total to 7 yesterday.

If we can safely ignore the international arrivals on the assumption that they will be contained by quarantine, then we appear to be on track to move into Level 3 with no more than one new internally generated case per day.

That being the case, the next question is can we maintain a very low number of cases per day in Levels 3, 2 and 1?  I guess the answer is that we can’t be sure, but indications are that the government is on track to be able to achieve this, with the continued co-operation of Kiwis, of course.

If we then assume that we move to Level 2 shortly after the next Cabinet meeting on 11th May, then we could be at Level 2 by mid to late-May, allowing most businesses to restart, albeit with some limitations, subject to demand, and with restrictions on travel.

Once we get to Level 2, then electricity demand could come back to relatively normal levels over winter.  Currently, most of the very large industrial electricity users, with the exception of the Tiwai aluminium smelter, are operating at very low levels of demand, and these will start to come back under Level 3.  Some of the most affected sectors of the economy, including travel and tourism, hospitality, accommodation, however, are going to come back at varying but generally much slower rates. 

So, in the longer term, after winter, demand is likely to be down on last year:  the question is by how much?

There is a suggestion that we should have pursued a less restrictive strategy like that of Australia’s.  Just to add my “two cents” worth to this debate, I have noticed that Australia’s daily case rate has stopped falling and appears to have plateaued at around 45 new cases per day, whereas New Zealand’s daily case rate is still falling and about to approach zero.