top of page
  • Writer's pictureGeoff Russell

Gambling on the grid

Suppose you are trying to shift Australia’s 30 million tonne grain harvest and somebody has decreed that you couldn’t use anything other than utes. I’m sure somebody would suggest some market to manage the process, probably more than one market. And those who believe markets guarantee efficiency might believe that the resulting mess would make for efficient use of the utes. But the real way to solve the problem efficiently isn’t with utes; regardless of the elegance of any tower of markets built for the task. The way to move the grain efficiently is with huge trucks or trains or ships, depending on geography. No amount of clever market design can make up for the inefficiency of using inappropriate technology. That's the take-home message of this post, but it will take a while to spell it all out.

My last post drew attention to the grid, as an entity in its own right worthy of serious thought, in contrast to merely thinking about the generation of electricity.

Much of what is happening in Australia’s main grid, the National Electricity Market (NEM), is happening elsewhere in the world, but I’ll focus on the NEM.

I did a very rough estimate in that last post of $25 billion as the cost of upgrading the grid transmission in Australia specifically for wind and solar; based on the 10,000 kms that AEMO reckon is planning. The Australian Energy Market Operator (AEMO) Integrated System Plan (ISP) reckons they will only cost $12.8 billion and pay for themselves in benefits. What kind? “Scenario weighted market benefits” is what they reckon in lieu of listing anything concrete. What they seem to mean is that there are other scenarios which would cost $28 billion more than their “Optimal Development Path”, so the $12.5 billion must be saving $28 billion. That sounds like “Buying a Tesla Model Y for $54,000 will save you $40,000, because it isn’t a Model S, which costs $94,000.” I could save even more by not buying a Lamborghini Aventador.

Is this cost estimate credible? In the US, the recent Infrastructure Act is also heavily investing in grid rebuilding. It’s aiming to spend US$1.2 trillion on its upgrades. How does that compare? The US and Australia are both geographically spread out countries, but the population of the US is about 13 times as great. To match that US per-person grid expenditure, we’d need to spend AU$130 billion. Which is rather more than $12 billion.

You can see why I mostly avoid talking about money. It’s way too hard to distinguish trusted financial estimates from the musing of a Sam Bankman-Fried; who was a trusted financial guru … until he wasn’t.

I also remember the cost estimates for Project Energy Connect, the new SA-NSW interconnector, which started at $1.5 billion and ended at $2.4 billion; and it’s not done yet. Or Snowy Hydro which was supposed to cost $2 billion and looks like it will take 10 years and not 4, and cost $10 billion; and it's not done yet. So forgive me for being incredibly sceptical about the ISP cost estimate. But it’s not the cost overruns that worry me so much as the incredible confidence with which profound engineering problems are glossed over in announcing this kind of massive infrastructure project. It’s like: “Renewable intermittency? Don’t you worry about that, we’ll just modify the grid and add some storage.

And then there’s the practical matter of actually building the extra transmission. It’s easy to draw lines on maps representing an ideal pattern of transmission lines. Modern mapping software makes it super easy to turn that into a terrific Power Point presentation which will enthuse politicians and have people cheering in the aisles. Building such transmission lines is another thing. It’s a bit like building a nuclear reactor, except you don’t only have to deal with a local population around a small site, you have to deal with the local population in every single kilometre of the proposed path. Legal action is already underway against the transmission lines for Snowy 2.0. Such actions could transform Snowy 2.0 into not just a late, expensive project, but one unable to deliver its capacity to where it is needed.

Even small transmission projects can be fiercely opposed … because nothing of this scale is small when it’s next door. Opposition to transmission lines for the Bulgana project a little north west of Melbourne is also ferocious; even more so being in marginal electorates!

In short, nobody has done what the ISP is proposing, nobody knows how to do it, the list of problems is long, and sometimes complex, but they reckon they know what it will cost. Really?

Time to discuss more concrete issues.

Moving on, digging deeper

This post will deal with mainly with capacity and congestion, but I’ll also add a little bit about the obscure, but perhaps a canary-in-the-coal-mine topic of sub-synchronous oscillations.

Capacity and congestion are the subject of a considerable amount of work at present as grid managers try to juggle what happens when you have nothing up your sleeves but bad choices.

First, some quick definitions before we get started.

  1. Capacity. It doesn’t matter how many gigawatts of solar panels you have when the sun goes down. Wind power is more complex. How many gigawatts (GWs) of turbines do you need to guarantee that at least 30 GW will be available on a given day? Keeping the lights on is all about what you can be absolutely sure of. It’s guaranteed capacity that has sent Europe scrambling for gas, Liquified Natural Gas (LNG), building regasification plants, re-opening coal and nuclear plants, signing new nuclear build deals, opening nuclear power training courses and so on. The Energy Security Board (ESB) and its associated experts tried to design a capacity mechanism to guard against the capacity problems in Australia in a mostly, or all, renewable grid. But somebody persuaded Australia’s energy ministers to throw this out. If you think Donald Trump and the Brexiteers have a monopoly on rejecting expert advice, then think again. AEMO and the gang (ESB, AEMC) have been doing their level best to work with what they’ve been instructed to do, but having yet another plank stripped from the sinking ship by a bunch of politicians must be supremely frustrating. Nevertheless, I have a sneaking suspicion that some of them are loving it for reasons that will become obvious.

  2. Congestion is actually a couple of related concepts. The first is a local issue where you have too many generators in a region for the available transmission capacity. According to the ESB, this should be considered as normal, not a bug. Perhaps the ESB should have been employed decades ago to sell traffic jams to commuters and we wouldn’t have so many concrete overpasses. Rising congestion costs are, the ESB points out comfortingly, a global experience. That’s a relief, misery loves company. The second meaning of congestion is more global. It’s not about where the solar and wind farms are – there are just too many of them. They are trying to sell either too much energy or trying to sell it at a time nobody wants it.

  3. If “sub-synchronous” oscillations seems scary, then think back to your bicycling days. Did you ever experience front wheel wobble? If you did, you’d remember. You probably fell off. It starts as a vibration in the wheel and handlebars and then, if you don’t respond fast enough, the wheel wobbles move from small to large, and then, quicker than you can blink, you are on the road. Modern racing bikes are extraordinarily well-engineered and such events are rare, but it can still happen given the right combination of wind, speed and road surface. In a power line, the voltage or current can oscillate in damaging ways also. This is over and above the normal up and down of the normal alternating current waveform.

Let’s get going.


Consider the following week’s worth of electricity output on the NEM using the OpenNem tool:

NEM week ending Feb 1, 2023

The grey vertical bands run from 10pm to 7am; (9 hours). You can see that there isn’t much solar power around during that time – just a little in the morning. How much electricity is used during that period? You can get a rough estimate by eye; just multiply the 20,000 megawatts on the left by 9 hours; giving 180,000 megawatt-hours of electrical energy; or 180 gigawatt-hours. This works because energy is power multiplied by time. Five thousand watts for 5 hours is 25 thousand watt-hours… 25 kWh. Similarly, a gigawatt-hour (GWh) of energy is the amount of energy delivered by one gigawatt of power over the course of an hour; power multiplied by time. It gets written with a hyphen before the time; watt-hour, megawatt-hour, sorry to labour the point, but I see journalists screw this up on a daily basis. If they don't know the difference between power and energy, why are they writing about this stuff?

What’s the range of wind output power during those 9 hours?

That’s a bit harder to estimate by just looking at the image, because the band is so thin and sloping.

So here’s a different graph, also from OpenNem. It shows daily wind power output for every day in 2022.

Wind output on the NEM 2022

On this image, you can see the minimum wind output and the maximum, even if you can’t tell the exact numbers. If you go to the website you can interactively find the maximum and minimum. The maximum wind contribution was 147 GWh on 4th August while the minimum was 15.6 GWh on 8th August. Given we had about 9.2 GW of wind on the grid in October, you can see that the maximum was about 66% of the theoretical maximum (where all of the 9.2 gigawatts of wind were running at full power for a day) and the minimum about 7% of the theoretical maximum.

What’s the point of this?

According to the AEMO ISP, the NEM (which doesn’t cover WA or the NT) will have 141 GW of utility-scale wind and solar by 2050; reasonably evenly split, say 70 GW of each. But the NEM will need to be generating twice the electricity in 2050, because of the electrification of more things. So we could be needing 360 GWh (2 x 180) overnight.

We can ignore all the solar farms and panels and we can estimate that the minimum amount of wind output will be about 7% of maximum.

This means that 70 GW of wind will provide, at worst, about 7% of its potential maximum over a 24-hour period. Or about 9/24ths of this during a 9-hour period. But it could be worse; we are estimating a 9-hour minimum using a 24-hour minimum. In any event, on a still night, you could be getting only about 44 GWh of the required 360 GWh from utility wind and solar plants. And what if we get more wind power than anticipated by 2050? Say 100 GW? Will that fix the problem? It would give us, at worst, about 63 GWh instead of 44 GWh.

Snowy 2.0 stores plenty of energy, but can only deliver 2 GW of power. Meaning that over a 9-hour period it can only contribute 9 x 2 = 18 GWh of energy. There will also be 10 GW of gas, which could contribute 90 GWh.

Let’s summarise that, we could be needing 360 GWh overnight in 2050. But on a still night, we will only have 44 to 63 GWh (wind) + 18 GWh (Snowy) + 90 GWh (Gas) = 152 to 171 GWh. And then there are batteries. The ISP only predicts about 100 GWh of deep storage (Figure 23 in the ISP). There is still a considerable shortfall.

And we are generously assuming that everything is where we need it and not impeded by transmission or congestion issues.

My calculations above were simple. If anybody knows where in the ISP the missing overnight electricity will come from, then by all means let me know. But it’s not clear to me how the ISP can possibly guarantee to keep the lights on in 2050; even if everything runs to plan. And what if it doesn’t?

The ISP also has no deep storage closer to Sydney in 2050 than Snowy 2.0, so we can expect grid congestion to be a factor in getting any available stored electricity to Sydney.

Can we build enough deep storage to handle not just one still night, but a succession of them; perhaps with cloudy days? Can we build them in the right places? Can we ensure they are charged when they are needed rather than when it is profitable for their owners? These are all critical questions which I’ll say more about below.

On the other hand, nuclear-based grids are just ordinary grids which everybody knows how to build and run. We already have a couple: the NEM in the east and the WEM in the west.

As you can see from the following graph, nuclear power is already huge in the current European grid(s) and, in the absence of gas and coal, is well set to handle the capacity needs; the stuff which can keep the lights on. In the absence of a total European-wide brain blackout, nuclear will continue to hold up the sky over Europe in the absence of coal and gas.

Electricity by source in Europe 2021

The complexity of the storage problem isn’t just building the storage, but building it in the right places. This brings us to congestion.


I said earlier that this was too many generators for the available transmission lines. You can get a better idea of what’s happening by thinking of buses (or trains). Each type of bus has a maximum seating capacity. Count all the full buses during peak hour in the morning and you have a measure of congestion. A power line also has a maximum power capacity and generators are chosen so as not to exceed that capacity. The capacity is an example of a constraint; a rule you have to stick to. There are others. In the transport analogy, roads also have a capacity. It’s measured in vehicles per hour but will depend on the vehicle mix. The capacity constraints in a grid are less familiar, but are mostly to do with things overheating; like transformers and switches. Most of the boxes and can-like things you see bolted to electricity poles are transformers.

The critical feature of a transmission line is its maximum power capacity. Network transmission maps typically have a kilovolt rating on the lines, but that’s just the operating voltage. NEM constraints are in MW. When that power is reached, engineers say the constraint is binding or perhaps tight.

AEMO publishes data on congestion every year. It lists all the constraints applied during the year and the number of hours they were binding. In 2021 (the latest data) there were about 1,000 constraints listed. Some 595 constraints were binding in 2021 that didn't bind the year before. 245 of these were binding for over 24 hours in 2021. In short, congestion is rising on this and other measures.

Thinking back to the bus analogy. When the bus you want is full, you have to wait for the next one. Except if you live conveniently close to an alternative route. In which case you catch a different bus. It’s the same with the grid – sometimes there are alternatives. The big complication with the grid comes when your alternative cable runs to somewhere where the price of electricity happens to be different from where you are. Suppose you are in North West Victoria and you can’t sell your electricity in Victoria because the grid is locally saturated, you may find that the interconnector to SA has both capacity and demand. Your wind farm can deliver its electricity to SA.

But the flow of both electricity and, critically, money gets tricky in such situations. On the 2nd and 3rd of February in South Australia, we had a bucket load of wind power – far too much for us to use. But none of it went to Victoria! Why not? Perhaps it was congestion, or perhaps some tricky little detail of the rules pertaining to money. Either way, the electricity was dumped.

One of the problems created by renewables is the weakening of the constraint system. I mentioned this in a previous post. All current constraints have a particular mathematical form that allows a computer algorithm to tune the grid every 5 minutes to the prevailing physical and market conditions. Some new constraints, as a result of intermittency, don’t fit into that scheme and require new thinking and systems. This is at best, risky, and at worst, destabilising.

I linked earlier to an ESB webpage on “Transmission access”. The ESB published a “Directions” paper in November 2022 summarising consultations and work on Transmission problems. Most of this report is about congestion; but not how to fix it, but how to monetise it.

Cables and congestion

To understand congestion in the grid, you first need to understand light bulbs. Not the modern LED bulbs, but the old fashioned incandescent bulbs.

Pass enough current through a wire and the wire will eventually get hot enough to glow. Think toasters if you are too young to remember incandescent globes. All that heat was wasted energy. We were only ever after the 10% that was the glow.

All transmission lines waste energy for precisely this reason. There are other reasons also, but this is the big one; heat. There are all manner of trade-offs involved. Lines that use all kinds of tricks to reduce waste are more expensive than just using “simple” wires of copper and aluminium.

It makes plenty of sense to reduce wasted energy when you are making that energy by burning stuff, be it trees, gas or coal. But when the energy is zero carbon, then what exactly are you saving? You might save on generating equipment, be it wind turbines, solar panels or nuclear plants, but eliminating waste doesn’t come cheap and sometimes it costs more to fix a problem than it is worth. When you have clean energy, knee-jerk judgements about waste aren’t so easy. The Swedes use far more electricity than we do in Australia, but it’s almost all clean, … being either hydro or nuclear. So who cares how much they use?

Now, getting back to those power lines.

How do you determine the power being transmitted in a power cable? You multiply the voltage by the current. So if you double the voltage, you can halve the current and still transmit the same power. This is kind of like magic because it’s the current which makes the cable hot, not the voltage. This is why long distance cables use high voltage; to reduce the current. Higher is better, except that if you raise the voltage too high then the electricity can jump very big gaps; meaning you need to put the wires higher to avoid fires, for example.

If you double the current in the wire how much electricity gets wasted as heat? Is it double? That would be good, but no, it’s 4 times. And if you triple the current, then the waste energy rises by 8 times; and so on.

All of this means a transmission line has a capacity beyond which heat losses rise rapidly. It’s not quite like the fixed maximum capacity of a water pipe, but close.

Consider the following super simple example. We have a couple of solar farms on a transmission line connecting to some place wanting electricity. Keeping things super simple, let’s suppose the solar farms are 100 MW each and the demand is 150 MW for the next hour.

Assume also that the transmission line can only handle a maximum of 100 MW. There is a second transmission line and we can just assume it will handle any shortfall. Perhaps it can handle 500 MW and has a coal plant attached somewhere. We’ll assume that the solar and battery plants can undercut any bids by that coal plant to supply power. But the coal plant can run at night. Hands up anybody who thinks the coal plant will want to stay in business just to supply overnight power, especially if it is competing with wind.

Sample network

When both solar farms are producing at maximum, then we have 200 MW. That’s too much, not only for the transmission line, but is 50 MW above the demand.

It would be good to use the excess to charge one of the batteries, but which one? We can’t charge the battery closest to the demand because the line capacity will be full meeting the demand. So we’d have to supply 100 MW with electricity from S2 and direct S1’s output to the battery B1. How much will the battery owner want to pay for that electricity, knowing that S1 has no alternative use for it?

But what if battery B2 happens to be fully charged? Then it can supply 100 W for as long as it has capacity, which would leave the solar farms in a low-bidding war over who will charge battery B1. Note that batteries are always characterised by two numbers, maximum power and the length of time they can supply that. So a battery might be characterised as 100 MW/150 MWh. The first number is maximum power and the second is energy; the amount the battery can store, 1.5 hours worth at maximum power. You could also supply 75 MW for 2 hours with such a battery. I didn’t specify any characteristics for the batteries in the diagram, but let’s assume they are 100 MW/150 MWh.

How can a simple diagram have so many options?

If one company owned all four plants, then they could optimally plan to maximise the use of the equipment. Optimisation problems like this are ubiquitous in large corporations owning many assets. But with multiple owners in competition for limited capacity on the grid and producing far more electricity than is required, the result is more like dumping some UFC fighters in a cage with a barrel of cash.

Lawyers will pick over every line of every rule in the book, looking to maximise both employer gains and billable hours. The book in this case being the 1,821 pages of the Australian Electricity Rules.

Optimal use of resources isn’t a goal and there’s absolutely no reason to expect it.

The ESB points out repeatedly that the way the current market and grid works provides plenty of incentives to put things in precisely the wrong places. It allows a solar or wind farm to gazump other farms by judicious positioning which will give it a market advantage while increasing congestion at the same time. As the ESB puts it:

New generation and storage will continue to locate and operate in ways that are inconsistent with minimising total system costs. One likely consequence is elevated congestion, which means electricity cannot be dispatched to meet demand at the lowest possible cost. In turn, this will drive the requirement for more transmission investment to alleviate the congestion, which would not have been needed if the investment and operation of generation and storage had been efficient. The cost of this additional transmission investment is borne by consumers.

A renewable grid with its cluster of small, but high-valued, sources and batteries is like a bait ball to a bunch of great whites; aka lawyers. They will fight long and hard, at whatever is the going hourly rate, to preserve the rights of the solar or wind farm that who is paying them to maximise their income, and if it fucks it up for everybody else, then that creates even more work. All they are required to do is stay within the current Australian Electricity Rules.

Changing these rules is a bugger of a process, meaning more jobs.

The ESB Transmission Access Directions report is a gold mine for lawyers and economists. Consider the following diagram which describes the various design choices for their proposed market to reduce congestion problems:

Congestion moneitisation methodology

There are 30 boxes under the various categories of design choices (aka job opportunities). The possible permutations make powerball look trivial.

The “Congestion relief market” (CRM) is the main financial instrument to try and send signals to investors to improve location choices for equipment. In addition to this market, the ESB envisages other tradable things which can monetise the lack of actual planning methodology. Remember the old nugget about Australians being willing to bet on anything, even flies crawling up a wall? The ESB sees the world through this kind of lens. Anything can be bet on (aka traded), and hence monetised.

Here’s an analogy. Think about a call centre; everybody knows and loves them. You may want to have your call answered quickly, but adding enough capacity (meaning people on the other ends of the phone) to make this happen would be very costly, with most of the new staff sitting around doing nothing for considerable periods. So it is with transmission lines. If you have 5 solar farms in an area, providing transmission infrastructure to handle all of their output all of the time would be incredibly costly; particularly if there is nothing to use those very expensive transmission lines after sunset.

The ESB view is different, why bother to solve the problem when you can monetise it? Their approach to both problems is trading. Why not set up call centre queue trading schemes? When the call centre tells me I have a 1-minute wait, then set up a system to enable me to sell that to somebody with a 10-minute wait. And then there’s the call centre futures market, where I can bet on the time it will take for my call to be answered. Has anybody tried to call Telstra or Medicare lately?

Congestion and the big picture

I remarked above that ESB reckon congestion will be a fact of life under high renewable penetration. They see it as a feature rather than a bug; it’s yet another thing which can be monetised with a market to make jobs for wanna-be Enron boys and girls. It’s really only a matter of time before we can add GridBet to SportsBet and the TAB.

The ISP predicts a NEM supplying double the current NEM output of 180 GWh annually. Remember this doesn’t cover the whole of Australia, which uses about 270 GWh. 180 GWh could be supplied by 20 GW running continuously, but that wouldn’t handle the peak demand which is closer to 30 GW these days. The plan is for 140 GW of wind and solar, meaning an overbuild of almost 5 times, plus an additional 69 GW of rooftop photovoltaic (PV). What happens in the middle of a hot windy day? We have something like 150-200 GW of power trying to supply 30 GW of demand or charge batteries. Of course, that’s a recipe for massive congestion and simply throwing away electricity.

The ESB estimates that over 20 percent of the electricity generated by wind and solar farms will be unused in 2050 – about 80 terawatt-hours. There are three components to this. Some energy will be thrown out purely because the wind or solar farm won’t want to sell it at the price available. A second component will be thrown out because there is no grid capacity to get it to where it is wanted. The last component will simply be a shortage of demand. When you have over 200 GW of capacity feeding a system that wants only 30 GW for a large part of the day, then throwing energy out is inevitable.

All these electricity losses are on top of the expanded losses by having generators so far from consumers. The unused fraction is, by the ESB’s own admission, an underestimate, being based on everybody building exactly where the ISP calculates is optimal. As noted above, this isn’t happening.

There will be a proliferation of lawyers fighting over the details of the rules trying to get their paymaster’s megawatt-hours accepted over somebody else’s.

As the ESB says:

In some cases, generators are connecting in locations where, a lot of the time, they are not adding new renewable energy to the power system. Instead, they are displacing existing renewable generators.

This is the gazumping we mentioned earlier. It doesn’t have to be deliberately malevolent. The ESB spends some time describing the kinds of information about congestion available to investors seeking to locate a new plant, and it is clearly insufficient. But again, this isn’t deliberately malevolent either. Congestion in the real world is far more dynamic than my simplified example and the best you may be able to give investors is probability distributions of risk.

Does all this smell like unproductive industrial churn? Plenty of money changing hands but not much actual value creation.

Consider the household case. A friend of mine has a rooftop solar system with a failed inverter. The original installer has gone broke and no-one else will replace the inverter, they all insist she upgrades. This is an old story with consumer-grade junk, stuff that is made for selling rather than for operating; and never for fixing.

You’d expect that solar and wind farms worth hundreds of millions of dollars and forming part of our national infrastructure would be different. When a new coffee shop opens halfway between an existing coffee shop and the office block providing many of the customers, the first coffee shop will lose custom. That’s unfortunate for the first coffee shop; but it’s not of national import. But when a $400 million dollar solar farm can be screwed over by another opening up and causing congestion, the significance spreads far wider. That money will have come from all manner of investors, including perhaps Super funds and Banks.

Sub-synchronous oscillations

This last section is fascinating if a little obscure. I gave the analogy of bicycle wheel wobble in the introduction. Many industrial systems can suffer or fail completely as a result of oscillations. There are various kinds, and they can bring down bridges, see here and here.

AEMO has been investigating why voltage oscillations have been appearing in the West Murray Zone of the grid. They started investigating in August 2020 and the most recent report was in October 2022. They still don’t know the cause.

This isn’t an entirely new problem in electrical grids, but one made much harder to analyse as grids become more complex. A recent paper discussed the particular problems of wind farms connecting to weak grids:

Due to the high proportion of renewables and power electronics, power systems are facing emerging power stability challenges that are different from classical stability issues

In a previous post, I pointed to various problems that I think may end up being really hard to solve. For me, it was the inability to model component failure with enough accuracy and speed to plan. Is the oscillation problem small or big? Here’s one opinion from somebody far more qualified than me to say.

According to a 2020 system strength workshop held by Australian Energy Market Operator (AEMO), weak grid associated stability challenges are viewed as the most significant challenges to higher IBR [Inverter Based Resources] penetrations.

Summary and conclusion

The transformation of our electrical grids has been largely forgotten in the debate about what kind of clean electricity future we should have.

It’s pretty clear from the long list of problems in the AEMO Renewable Integration Study: Stage 1 (RIS) (Stage 1) that the engineers understand the enormity of the task, but for many others, working on huge projects is unfamiliar territory, and the risks of failure don’t seem real.

Remember how the world changed when Covid-19 arrived? Experts had been warning about such things for decades. But if you haven’t experienced such events, they don’t feel likely. For many it was akin to worrying about an alien invasion.

The ISP isn’t a plan for a few more transmission lines, but a plan for a fundamentally different grid.

A failure on the scale of the $250 million ASX CHESS system isn’t something familiar to most people. And that’s actually a small project in comparison to the many much bigger tasks involved in the redesign and rebuilding of the grid. The size and complexity of a task isn’t simply a function of the price. Plenty of projects with a $250 million price tag are relatively trivial. The complexity of grid transformation is shown not only by the list of subtle and complex problems highlighted in the RIS but by the simple fact that people have been working on it for about 15 years and the list of problems is growing rather than shrinking.

How do we tell the difference between hubris and the bravery required to tackle big tasks?

The public has been sold on renewables despite their failures during the oil crises of the 1970s. Most are simply too young to remember that debacle. Solar panels on the roof of the White House. Fields of now rusting wind turbines. And it’s happening again. In contrast, the French decarbonised much of their grid with nuclear power in about 15 years; by using a tech that didn’t break the grid.

Renewables have done little in any country over the past 20, despite massive support in many places. How can a technology achieve so little over such a long time period and still have support? That’s actually easy. It’s been a perfect storm. The Chinese have made panels cheap, and politicians have loved being able to ignore personal conflicts of interest and legislate for big feed-in tariffs which made it profitable for both climate hawks and climate deniers to hop on the bandwagon. There’s an old saying: “For every complex problem there’s a clear, simple, plausible solution which is wrong”.

Solar has captured people’s imagination by fraudulently implying that the climate problem is simply fixed if we all pitch in and put a bit of tech on our roofs. Wind similarly.

But all is not lost … yet.

Despite the US commitment to modernise it’s grid, mentioned earlier, the US is finally starting to ramp up support for new nuclear plants. TerraPower, one of Bill Gates investments, is looking to build 5 reactors. Where? On the site of coal plants; where they can keep all the grid investments. The same will be true for hundreds of coal sites in the US. The US Department of Energy found it could raise the US nuclear capacity from 95 GW to 350 GW. Again this would mean keeping all those grid investments. The Chinese have been planning such a move also and recently connected the first reactor to their grid deliberately designed for the task. In the US it may well be the Holtec Small Modular Reactor reactor which will do the job.

Unfortunately, the US has a solid track record for totally botching nuclear reactor builds. It has a special Government organisation tasked with slowing nuclear progress, the Nuclear Regulatory Commission. It’s been extremely successful for over 40 years. While the Japanese built 60 reactors with a median build time of just 3.8 years, the US median build time was about 8 years. Can the US reform or abolish the Nuclear Regulatory Commission and get some sanity into their processes? Who knows. In any event, it isn’t a surprise that the US is finally having a bet both ways, renewables plus nuclear.

Australia is increasingly isolated in thinking it can succeed where everybody else is failing. Japan has recommitted to nuclear power, as has France, the UK, South Korea, the US, not to mention China, India, and Russia. Our current opposition to nuclear power fits in well with our gambling culture and preference for markets and magic rather than engineering as the means to solve problems.

Appendix: Rollout rates of renewables vs fossil fuels and nuclear

For an explanation of the following chart comparing nuclear and renewable rollout rates, see here.

Fastest 20 year electricity rollout rates


bottom of page