Texas has experienced an unusually severe winter storm and the state’s electricity infrastructure has not been able to cope. This situation has prompted a round of climate-based political recriminations with conservatives blaming an over-reliance on renewables and progressives pointing the finger at gas plants and a history of light-touch regulation. How much of this political back and forth has an element of truth and what lessons can Australia take from events in Texas?
First, it is necessary to understand that Texas is different to the rest of the US — it has its own power grid that is essentially separate from both the Western and Eastern Interconnects that cover the other 47 contiguous states. There are various historical and political reasons for this, including the fact that Texas power companies were the first to recognise the value of working together under what is now known as an Independent System Operator (ISO). This happened during the Second World War and was designed to ensure a reliable power supply to critical wartime industries. In more recent times, the fact that a Texas-only system, now run by the Energy Reliability Council of Texas (ERCOT), can operate outside the purview of the Federal Energy Regulatory Commission (FERC) has had significant political attraction for independently minded Texas administrations.
Comparison of Texas electricity generation with Australia
The table below summarises some of ERCOT’s key operating parameters and compares these with the National Energy Market (NEM) in Australia. Note that ERCOT covers about 75 per cent of Texas by area and 90 per cent by population while the NEM covers all of Australia except for Western Australia and the Northern Territory. The NEM system operator performing a similar function to ERCOT is the Australian Energy Market Operator (AEMO).
|
ERCOT[^1] |
NEM/AEMO[^2] |
Annual Generation (TWhours) |
380 |
202 |
Peak Demand (GW) |
75 |
36 |
Installed capacity (GW) |
82 |
49 |
Distribution network (km of power lines) |
74,000 |
40,000 |
Total area (square kms) |
522,000 |
3,675,000 |
Population served (millions) |
26 |
22.5 |
Generation source (% of annual output) |
|
|
Coal |
18.0 |
66.1 |
Gas |
45.7 |
7.4 |
Nuclear |
10.9 |
|
Wind |
22.8 |
9.7 |
Solar |
|
9.6 |
Hydro |
2.3 |
7.0 |
Residential power price (US$/kWh) |
0.085[^3] |
0.15-0.22[^4] |
ERCOT is roughly twice the size of the NEM in terms of total power produced, peak demand and installed capacity. While the population served is roughly equal, Texas has a much higher industrial base, attracted in part by low power prices.
Texas, like many US states, has cheaper power than Australia, though this differential has closed in recent years. Relative to the rest of the United States, Texas has relatively low-priced electricity, though not as cheap as Wyoming which has cheap, local coal or Washington which has Columbia River hydropower. Texas does have considerably lower power prices than big state rivals California and New York.
As to who and what was to blame, the two main finger-pointing arguments were either that wind generation, which in some places dropped to zero as turbines froze, was inherently unreliable or that gas extraction, distribution and the gas-fired power plants themselves had not been equipped to handle cold that would have been routine in many parts of the country.
Renewables supply just over 20 per cent of Texas generation. This is predominantly based around wind power, with Texas having one of the largest wind fleets in the world but very little solar. Australia, on the other hand, is only slightly less reliant on renewables overall but with roughly equal contributions from both wind and solar. Turning to baseload generation, the NEM still relies on coal, while gas supported by sizeable contributions from coal and nuclear provides ERCOT with a more diverse generation mix than Australia. One factor the NEM and Texas share is that they are generation islands — an inescapable fact for Australia but a choice for Texas. This means they both need to be self-sufficient in terms of generation capability — able to meet their own power demand needs. This is in sharp contrast with other US states and indeed many other nations who regularly buy and sell power with neighbouring jurisdictions.
What went wrong in Texas?
In simple terms, supply was not available to meet demand — the nightmare scenario for utilities and power regulators. The demand at the start of the cold snap was about 69 GW, higher than normal for winter but well below summer peaks. Unfortunately, about half of the installed capacity of 82 GW failed under the extreme weather (for Texas) resulting in a 30 per cent supply reduction. Rolling blackouts were mandated to ensure that the available power could be despatched in a stable fashion.
Australians and Texans have both come to expect a high level of power reliability so the recent situation in Texas, which led directly to a number of fatalities, can quite rightly be called catastrophic. As to who and what was to blame, the two main finger-pointing arguments were either that wind generation, which in some places dropped to zero as turbines froze, was inherently unreliable or that gas extraction, distribution and the gas-fired power plants themselves had not been equipped to handle cold that would have been routine in many parts of the country. ERCOT’s announcements highlight the latter over the former. As is typically the case in these situations, there were compounding factors elsewhere in the system with both coal and nuclear plants either impacted by the weather or offline due to planned outages.
Contingency planning for extreme events
When the dust (or perhaps the snow) settles one suspects the focus will shift to ERCOT’s operating philosophy which, partly as a result of being outside FERC oversight, emphasised low costs over resilience. To ensure a power grid can operate under rare but extreme conditions takes planning and money — money which is not be needed in normal conditions and can easily be labelled “gold plating” especially if private utility companies can pass the costs onto a paying public. After a cold spell impacted power supply in a number of South Western states in 2011, FERC issued a report recommending a range of corrective actions. It seems ERCOT deemed these recommendations neither mandatory nor necessary and that “weatherising” wind turbines and gas plant infrastructure would simply add to the cost of power. In February 2021 this judgement looks flawed.
Australia can probably choose to ignore the risk of prolonged -30 degree C temperatures but we do have exposure to floods, bushfires and other extreme climatic conditions that will likely be quietly reviewed in coming weeks.
Australia can probably choose to ignore the risk of prolonged -30 degree C temperatures but we do have exposure to floods, bushfires and other extreme climatic conditions that will likely be quietly reviewed in coming weeks. Australia, with federal and multiple state governments having oversight, should be better placed from a governance perspective to get the right trade-off between grid reliability and cost to consumers. Post Texas, this trade-off might shift a little to favour greater resilience.
One risk factor that, on balance, works in Australia’s favour is the size of the grid. The NEM generates half the power of ERCOT but covers an area seven times greater. As Australia converts to renewables, ensuring that generation is geographically distributed across the vast area available in Australia reduces both the natural intermittency of renewables and the impact of localised events.
Was an over-reliance on renewables (or gas) a critical issue?
Texas actually has a very diverse generating mix and in relative terms, ERCOT is less reliant on one particular technology than most grid systems. The specific suggestion by conservative advocates that unreliable wind power was a factor has effectively been rejected by ERCOT itself. However, while the implication behind the accusation is misguided, wind and solar generation are more variable than other forms of generation and the grid design needs to anticipate inevitable and potentially extended periods of low output. Hopefully, this observation promotes a more detailed and informed discussion about the need for Australia to respond to this reality and install suitable sized storage systems.
Australia — and one suspects Texas as well — has not fully engaged with the realities of a future increasingly based on wind and solar generation. Australia is moving inexorably in this direction, driven by the cost competitiveness of wind and solar, the limited lifespan of ageing coal plants and ongoing political pressure to decarbonise. Without a mechanism to ensure that shortage capacity keeps pace with growth in renewable generation we will be flirting with disaster.
Australia is moving inexorably in this direction, driven by the cost competitiveness of wind and solar, the limited lifespan of ageing coal plants and ongoing political pressure to decarbonise. Without a mechanism to ensure that shortage capacity keeps pace with growth in renewable generation we will be flirting with disaster.
In this context the structure of the energy market in Texas is relevant. Texas, like Australia, has an Energy Only Market (EOM), in simple terms, this means market participants are only paid for the energy (or services) they actually provide. The alternative is a Capacity Energy Market (CEM) which pays participants for having capacity available, even if it is not actually used. A CEM model is obviously more conducive to promoting storage systems that aren’t needed when wind and solar generation are plentiful and potentially won’t be fully utilised except in rare but extreme conditions. Few will invest in infrastructure with an uncertain and perhaps remote pay off but guaranteeing a steady income stream for being on standby and ready to support the grid when needed becomes more attractive.
Creating a capacity market in Australia is not a novel concept — this discussion is bubbling along among expert groups and mostly out of earshot of the general public. If we are lucky these experts will incorporate learnings from Texas and present a system redesign that works as required and avoids too much interference from political actors and the advocacy industry.
How much storage might we need?
Australia could also benefit from a serious discussion about what grid-scale storage means. Grid-scale storage to support generation based around renewables needs to be large — thousands of times bigger than South Australia’s much-hyped Telsa battery. This can be a sensitive topic given the potential for conservative groups to portray this as a new hidden cost for renewables. Climate activists on the other hand can struggle to get out of booster mode and talk honestly about the challenges and costs of transitioning to a carbon-free generation. It may come as a pleasant surprise to some but both ERCOT and AEMO do have updated and detailed roadmaps that incorporate lots of new renewables, major storage facilities as well a sharp decline in coal. A key hurdle will be political acceptance and in Australia, at least some of this is currently coming from state governments.
One of the subplots of a reliability discussion — particularly when presented as advocacy — are arguments that highlight outages with fossil fuel systems, suggesting it is unfair for renewables to be the only generation options labelled as unreliable. This is true up to a point but the difference with coal, gas and nuclear plants is firstly that downtime is not synchronous and secondly that it can be minimised through maintenance and regular upgrades. As we decarbonise this becomes a moot point for coal and perhaps gas but in the aftermath of a Texas-like event renewable activists can be reactive about a focus on reliability and the need for backup systems.
Can we avoid the Texas experience?
Suffering from a once in a lifetime event is bad luck and Texas has been unlucky. It has, however, been running close to the margin in terms of reserve margin and reliability requirements so while accusations that they got what they deserved are a bit pejorative, they are not without some substance. From a governance perspective, Australia and particularly the NEM has multi-government oversight which offers protection against groupthink and a reliance on a dominant operating philosophy. Hopefully, this means that when the entities running the NEM tell us a Texas-style event won’t happen here we can believe them. Even if true this doesn’t mean there aren’t some learnings we can extract.
The most important of these are around a broader public and political recognition that the trend that has seen Australian renewable generation grow from 7.5 per cent in 2015 to almost 20 per cent last year Australia will continue. In recognising the inevitability of continued growth, the key challenge is to design and implement systems that encourage and adequately reward investment in storage capacity, especially storage designed to support the grid during extended periods of reduced generation. A discussion on how we compare to Texas could help inform and guide this work but too much focus on slanted, advocacy heavy material could be a distraction.