Four Graphs and a Target

There are four graphs that I keep coming back to when thinking about current and future electricity systems. This is for the UK, but the logic is generally applicable to countries with similar strategies.

In the first graph it’s about what does the future look like compared to today. Only the far-left smallest single column is real, all the other columns are the outputs of various modelling efforts. The key common factor, as we increasingly electrify our society, is that they are all much much bigger in terms of the system generating capacity (GW) and the energy we require (TWh). The commonality between predictions helps us understand we a) need to generate a lot more electricity and b) the electricity system will be a lot bigger and a lot more complicated. We should also remember another common factor in that ‘all models are wrong, but some are useful’ (George Box). This is also just a snapshot of models in one year, look back in time (just a few years) and you can see that these models have also changed their predictions from year to year.

‘Prediction is difficult, especially when it is about the future’ (Mark Twain) and for energy particularly so.

As an aside, it's also interesting to reflect on bang for buck when it comes to the TWh we get from GW built. Today the UK gets for every GW on the system approximately 2.8TWh per annum. Under one of the proposed Govt scenarios this can drop to 2.2TWh per annum from 1GW in 2035 and by 2050 we may be getting to 1.8TWh or less. So this means building more infrastructure for less TWh. However, this investment decision may be balanced (with excellent system design and planning) against the benefits of a) a more efficient use of energy by avoiding burning things (reduction in energy losses between generation and use) pulling demand down, along with efficiency, and also b) reduction in the fuel we need to operate the system, primarily fossil, hence reducing external reliance and improving energy security and price volatility. However, it is hard to get away from the fact that electricity systems with greater levels of intermittent power generation, i.e. weather dependent, need a lot more physical stuff to run reliably.

Getting the generating mix right is key for affordability and security (system reliability and external reliances) as well as decarbonisation and sustainability.

So, we know the electricity system will be bigger and more complicated. But what pace do we need to build it at?

The second graph is about build rate and the targets set in policy statements. The graph used here is from last year and it shows the history of what we built in the UK in the past and when, if you look to the left of the big blue columns. Worth noting that in general we built one technology at a time (coal, nuclear, gas, solar and wind) not simultaneously and that these columns, showing GW connected to the grid, are A LOT smaller than the big blue columns. What the big blue columns show are the per annum amount of GW that would need to be connected to the grid and generating, also accounting for generating assets lost over this time, to meet 2035 and 2050 targets. These are very high. The UK connected 4.5GW last year and the per annum rate to 2035 thus went up to 15.5GW. This graph does not include the requirements of building the transmission system alongside the new generation technologies, this has to be done in parallel to connect geographical generation to demand otherwise serious issues occur pretty quickly – as we can already see today and why the detail, role and operation of the National Energy System Operator (NESO) is so important as are the plans required:  NESO will have to develop a useful Strategic Spatial Energy Plan (SSEP) and A Centralised Strategic Network Plan (CSNP).

These are ‘incredibly challenging goals’ (understatement alert). Bring the target forward to 2030 and the bars get higher. Don't have an actual plan to do it? Not looking great.

What provides some mitigation are the estimates on electricity demand, hence the scale of the system and amount of electricity we need per year, is intimately linked to the pace at which we electrify transport, domestic heating and industrial processes. If this slows down we need less TWh. However if we slow down electrification then we don't reduce our emissions from transport, domestic heating, industry etc. It's all interlinked.

One conclusion to draw from this picture is it is very unlikely that a single technology can be built at the pace required, and such an approach would also be likely detrimental to the energy trilemma we must constantly serve. Diversity is good to meet the needs of the trilemma for society's benefit. One might also conclude that the best way to meet such a step up is to go for it with a diverse mix of known technologies today that we know roughly where they will go and crack on, whilst developing new stuff and adapting appropriately. Rather than banking it all on the new stuff...

There is another potential pitfall with near term ‘targetry’ and that can drive us down making decisions that may only have near term benefits, whereas we should be much more interested in the longer-term system viability and the value gained from assets that would last for a very long time and ideally thinking well beyond 2050. This graph includes the contribution of power generating assets that will be certainly need to be replaced between now and 2050, and there are not many modelling scenarios valuing assets that go beyond 2050.

We should be wary of very near-term targets with regard to our electricity system and the behaviours that drives without a) a clear plan and b) a very good understanding of the ramifications of decisions. The UK has already done a lot on decarbonising our electricity supply and our emissions today are not a major addition to global emissions due to both the reduction in emissions but also due to scale.  If this rush means we don’t consider or prioritise power generating assets that take longer to build but will provide reliable power for long periods, over what might seem like quick fixes or untestable promises, then I think that will lead to a challenging situation for generations to come.

In the third graph we can see today what a low carbon electricity system looks like, and the good news is they already exist and for some pretty sizeable economies. This is October last year, but you can look at annual figures and see the same trends. The common factor between the low emission systems is some form of baseload or dispatchable low carbon power – geothermal, hydro or nuclear. The mix clearly depends on your country’s local resources and preferences. Replicating these systems would seem a practical approach, and in that replicating technology build out with fleet approaches and sustained policies.

However, not all countries are following that path and some are taking a different level of risk, generally revolving around making an assumption that long duration large scale energy storage will come good in the near term, either at grid scale and or extensively distributed. This also assumes that new infrastructure and the inefficiencies encountered in any energy conversions will be less risk and more value for money than a known technology path.

This approach is not low risk. In the UK the recent National Infrastructure Commission assessment was happy with the assumption that large scale hydrogen generation, storage and then burning hydrogen to produce electricity would be a more reliable and better value for money route than nuclear capacity exceeding 8GW in the UK. This is hard to prove today as we have very limited data on what these projects would cost and how long they would take at such a scale. Whilst it is not a bad idea to pursue large scale long duration energy storage I would not advocating putting all the eggs in one basket for something so critical to society.

If you want to explore electricity system emissions more then download electricity maps and have a play.

The final graph is a clincher for me when it comes to what we need to build at scale fast. I think it speaks for itself.

So, is there a lot to do to get to a low carbon reliable and secure electricity system to increasingly electrify society? Oh yes for sure this is not ‘in the bag’.

Can we do that with technologies already available to us and do we know what a low carbon reliable electricity system looks like? Yes. And whilst developing new technologies in parallel and adapting our plans whilst carrying out appropriate (engineering) risk assessments.

Is diversity of those technologies beneficial? Yes, for affordability, security and sustainability as well as balancing the amount of hardware we need and where.

Can we go faster? Definitely. And we could do with a good Strategic Spatial Energy Plan to do so, alongside practical reform.

Should we do that without due regard for the longevity of the electricity system for future generations and thinking beyond 2050 and do very near-term targets help? No.

Unlike archery, bringing this target closer just means we are less likely to hit it.

Our energy supply, and this will increasingly mean our electricity supply, is fundamental to both economic prosperity and societal health. We should be rigorous in our assessment and management of risk, make decisions on more than just the cost of electricity at source when generating, value diversity and think for the long term.

References: Atkins , George Box, Grant Chalmers , Mark Twain, Steve Hargreaves , Our World in Data , Fraser Steedman , Rozalia Buzova , Sarah Long CEng ChPP , Richard Beake , Electricity Maps , Dieter Helm