More Please…
Increasing renewable energy generation does not mean lower prices for the end consumer, despite there being zero fuel-costs. Why is this the case and what needs to happen to change this?
On Easter Monday, many of us enjoyed a quiet day away from the hustle and bustle of work. I was at the National Museum of Scotland with my son and niece, playing with the many science toys they have. The kids were fascinated to see how different objects and lifeforms appear under a microscope.
This was a change for me.
Usually on a workday, I spend my time looking at one of the many dashboards available these days, to see what is happening in the grid in the UK and elsewhere. Easter Monday was an extraordinary day for renewable energy generation in the UK. Wind and solar energy generated upwards of 16GW or ~60% of total demand.
While it was not a record, it was impressive and generated a lot of industry discussion on LinkedIn and the media more broadly. According to the Guardian, the record was on December 30th, 2022, when the wind sources alone were pumping in ~21GW or ~87% of total demand.
Figure 1: Generation and system demand in the U.K on April 10, 2023*
I am sure many people liked the fact that wind and solar dominated power generation on Easter Monday. In fact, at 4pm, Combined Cycle Gas Turbines (CCGTs) generated only about 6% of the UK’s power. Naturally, people are asking how we can have more of this, please!
Such days are in contrast with the popular belief that the electricity system operator (ESO) finds it hard to manage a power system with large levels of renewable power. The ESO and many in the industry, including myself in the past, have talked about lack of “inertia” or “system strength,” and have contemplated the role of “grid forming inverters,” and “grid scale battery storage”. It might appear to the educated layperson that it would be possible to have more renewable energy. Furthermore, since the cost of fuel is not a factor with renewables, that electricity should be available at a lower price.
To understand why not, one must ask whether this is the full picture? And as you might expect, it is not.
What you see in the Figure 1 is wholesale generation. This generation is a total sum of commitments made by the generators, retail suppliers and the system operator using many market and bilateral mechanisms, usually at least a day in advance. With advances in weather forecasts, it has now become possible to forecast the grid connected renewable generation to a reasonable degree. So, the system operator knows the generation mix for the following day. Hold that thought on the degree of accuracy because the degree of accuracy matters.
At any instance of time, the commitments made by the generators, suppliers and the system operator may only be fulfilled to a certain extent. The realities of weather, industrial, behavioural, social, cultural, sports and political activities converge to create a net difference between the generated and consumed power. This is called the power balance of the system. Despite it being the critical state of the electricity system upon which we all depend, there is no direct way of measuring the system balance. In AC power systems, the system frequency is the only effective proxy for this balance.
In the UK, the Balancing Mechanism is used to direct generation, or demand in some cases, to zero the system imbalance. As you can see from the Figure 2, wind energy sources stand at NEGATIVE ~856MW and Combined Circle Gas Turbines (CCGTs) at ~661MW. The negative sign signifies that the wind energy sources were asked to produce 856MW less than they were capable of at that instance.
Figure 2: The Balancing Market dispatch on April 10, 2023*
The ESO may do this for many reasons. They may have no network capacity to transport that power from where it is generated to where it is needed. Think of it as a truck hauling a large chunk of steel. While it may appear that there is more space on the truck, loading it anymore would cause damage. The ESO may also constrain the wind energy for system security.
System security can be threatened by any combination of a myriad of failures. Situations of a similar magnitude to the failure of large traditional generation can happen with the natural variation in available wind energy. It could be as simple as a small piece of equipment such as insulation failing, or as major as failure of a generating plant. Such a failure happened in England on August 19, 2019.
Back to the subject of accuracy in prediction. If the total wind generation is very small, a 10% uncertainty in the wind energy, considering all possible sources of variance such as prediction errors, randomness of windspeed or wind direction could be very manageable. On a day like Easter Monday, 10% deviation could be 1,140MW. This is larger than the system operator is obliged to secure the system for. On the record day of December 30th, 2022, 10% deviation would have been 2,100MW. The ESO might solve this in real-time either by paying wind farms to reduce their output or paying a thermal generator to be on hot-standby ready to pick up deviations in a matter of seconds.
This situation directly reflects on the financial outcome for the hour. For every MWh that the wind sources were asked not to generate, the system operator would pay them £82. At the same time, the system operator now must pay a thermal generator (like CCGT plants) to generate energy instead, at £158 per MWh. In comparison to the £56 per MWh of day ahead prices, the system operator is now forced to effectively pay £240 per MWh.
Figure 3: Comparison of day-ahead, intraday, and balancing prices*
A Moderate Wind Day
Figure 4: Generation and system demand in the U.K on January 23rd , 2023*
Let us compare the situation above with a regular workday and an evening peak on January 23rd, 2023. At 4PM, wind generation of ~4,523MW is about 10.5% of total demand and CCGT at 23,333MW is 54% of the demand. The energy at that time was very expensive with a day-ahead price of £212 per MWh.
Figure 5: The Balancing Market dispatch on January 23rd, 2023*
By stark contrast, the system operator is doing fewer balancing actions through the Balancing Mechanism. There is less risk overall measured by any technical metrics. In terms of balancing prices, the offer and bid price are both positive therefore to take a 1 MWh balancing action, the system operator would spend ~£246 per MWh. This is only 15% more than the day-ahead prices.
Figure 6: Comparison of day-ahead, intraday, and balancing prices on January 23rd, 2023*
The marginal cost of energy on a windy day with low overall system load was about £240 per MWh and the marginal cost of energy on a not so windy day was £246 per MWh. Expressed as the rate paid by end consumers, zero-fuel cost wind and solar energy sources hardly made a difference.
One sure shot way of improving this situation is by enabling a substantial system balancing cost reduction. To make this happen, we need market participants who better respond to price signals, networks that can flex and contract when needed. Until then, the cost of incorporating high levels of renewable energy is nearly the same as expensive gas. Any short-term win is only in a momentary reduction in emissions.
Watch this space! The solution may be just around the corner and then you may say “More Please!”
* Graphs plotted using Modo Energy’s Plotter Tool
