Getting to grips with intermittency05 Nov, 2012
Much has been written about the challenges of managing load intermittency arising from the growth of renewable capacity in European power markets. What is not so clear is a consensus on the scale and potential consequences of intermittency dynamics. At its simplest level this boils down to a capacity issue: how much intermittent renewable capacity is built and how much flexible capacity is available to manage it. Despite uncertainty around the absolute level of each, it is clear that intermittency will drive structural changes in generation and market price dynamics.
An OIES take on the issue
An Oxford Institute for Energy Studies (OIES) report titled “The Impact of Import Dependency and Wind Generation on UK Gas Demand and Security of Supply to 2025” (Ref 1), penned by Howard V Rogers last year, takes a detailed look at these issues in the UK context. The basic methodology applied by Rogers was as follows:
- Forecast the future levels of wind generation output based on wind speed data from a characteristic year (2009), turbine efficiency as a function of wind speed and future growth of installed capacity.
- Assess the impact on other installed capacity types based on a simple assumed merit order and extrapolated demand (based on characteristic day demand shapes).
Modelling boffins might argue that the above methodology was over simplistic and that a detailed fundamental market model is required to do the problem justice. However, the elegance of Rogers’ analysis is its simplicity and transparency. This neatly side steps the distraction of uncertainty over the detailed modelling methodology and input assumptions. More importantly it directly addresses the real issue with intermittency which is renewables load uncertainty in the very short term. By contrast, many long term fundamental models will be set-up to work on an expected outcome basis and work with average renewables generation (or derated capacity).
The OIES report offers some striking illustrations of the potential impact of intermittency. The charts below show the daily generation patterns of the different capacity types for 2015 and 2020 from Rogers’ analysis.
Chart 1. Wind penetration into daily generation stack (2015 & 2020)
Source: Howard V Rogers (OIES). Aug 2011. Ref 1 above.
The charts above clearly show the impact of increasing levels of intermittent wind (green) on the absolute level and variability of gas fired generation (red). This variability will be directly exported to the gas market as gas swing demand for power generation. The picture becomes becomes even more extreme when viewed at hourly granularity. The chart below shows the analysis for a charateristic week in 2020.
Chart 2 . Hourly wind penetration for characteristic week in 2020.
However it does raise some important questions as to the required levels of reserve during periods of high wind output and the increased probability of National Grid being required to constrain off wind capacity.
At a higher level, the key conclusions from Rogers’ report are as follows:
- The dynamics of the UK generation market will be dominated by the growing contribution of unpredictable wind generation over the second half of the decade.
- As the UK lacks significant reservoir and pumped storage hydro capacity, gas fired generation is the natural provider of short term flexibility required to service the intermittency of increasing levels on installed offshore and onshore wind capacity.
- Interconnectors may also provide transitory flexibility but their influence is likely to be limited due to the fact that the high and low pressure weather systems that drive wind speeds can span hundreds of miles across NW Europe creating similar flexibility requirements in neighbouring markets.
- Once installed wind capacity reaches approximately 28GW the ability of currently installed CCGT capacity to manage intermittency is exceeded. This represents a cap on the ability of the existing infrastructure to manage intermittency without arrangements to limit the output of capacity at times of high wind.
The first three conclusions offer a good, if uncontroversial, summary of the issues but the most eye catching conclusion is clearly the last. It poses an interesting question as to how soon the UK could reach this implied cap. The greatest uncertainty is clearly the assumed build rates for both onshore and offshore wind.
Uncertainty in wind build rates
The assumptions in the OIES analysis were largely based on data from DECC’s 2050 Pathways Analysis (published in 2010) which, consistent with its purpose, assumed a very aggressive path to decarbonisation, in large part delivered by a step change in renewables build rates. As illustrated in the chart below, previous Timera Energy analysis based on simple extrapolation of historical build rates suggests that getting anywhere near DECC’s assumed levels is highly unlikely unless there is a step change in policy support for renewables.
As the reality of the UK’s finances begins to bite and the inherent tension between austerity and decarbonisation policies comes to a head, DECC’s aggressive build assumptions feel increasingly optimistic. In contrast, Ofgem’s assumptions on wind build in its recently published Electricity Capacity Assessment report (Jul 2012) were closer to our more conservative build rate assumptions (included in the chart above for 2015 only as Ofgem’s study covered the period to 2016/17).
The other side of the equation is the level of available flexible capacity to manage intermittency. The Ofgem report highlighted the increasing awareness of the potential for a capacity crunch mid decade. In our view there is also a significant risk that a number of older CCGT plant are closed by the middle of the decade.
Impact on generation and market price dynamics
Irrespective of renewables build rate uncertainty, the OIES analysis highlights the structural changes in generation and market price dynamics that will increasingly dominate the UK market. These factors will be overlaid on weather sensitive demand and plant availability that have historically dominated prompt fundamentals, prices and volatility.
It is worth recapping what some of the broader implications are likely to be for the market:
- Increased load flow variability and transmission stress – intermittency is not just absolute load variability. The locational diversity of the intermittent capacity will also lead to increasingly dynamic load flow patterns which will make the system balancing tasks (including voltage, frequency and constraint management) of National Grid more challenging (as has clearly been demonstrated in Germany).
- Reduced and more uncertain gas plant load factors – Over time intermittent generation will change the structure of gas plant returns as they face reduced running hours and less certainty as to when they are likely to generate.
- Price cannibalisation – As has been the case in Germany with the increase in wind and solar capacity, prices will face downward pressure due to the increasing levels of low/zero marginal cost capacity. This price cannibalisation will impact both spot prices during low demand/high wind periods and forward prices from an higher expected level of wind generation.
- Increased prompt volatility – more dynamic and less predictable prompt fundamentals will be likely to increase volatility in both the power and gas markets. This will be a key value driver and market signal for assets in both markets providing flexibility in the prompt.
- Stronger coupling of gas and power market price dynamics – the dominant role gas plant will play in managing intermittency will likely strengthen the already high levels of correlation between gas and power prices and volatility.