What's in Your Electricity Bill : Part 2 - Constraint Payments

Constraint Payments

Summary:

• Constraint payments are paid when a plant is forced to run differently to schedule.
• These payments have risen since 2009 from €101m to €175m in 2014.
• Years with good wind output (high capacity factor) coincide with those years that had high constraint payments
• This is because wind is intermittent and must be taken by the grid when available (priority dispatch). This results in plant running differently to schedule more often than in a system with little or no wind.
• The cost of constraints in 2014 was roughly equal to the claimed fossil fuel savings by SEAI for 2012.
• Older plant have been replaced by modern efficient plant which require longer start up times but should result in more efficient use of fuel. But because wind is running in and out of the grid intermittently, the new modern plant are running less efficiently than they are designed to. This adds to constraint costs.

Constraint payments are a form of compensation for a plant when it runs differently to its scheduled production. I will explain this in further detail below but first of all I need to clear up a few things. There are different forms of constraint payments. The type of constraint payments discussed in this article are paid to conventional plant, mostly gas.  According to the Department of Energy, curtailment payments in the region of € 7.5 million were paid to wind farms in 2013. Curtailment and constraints are, confusingly, when talking about wind energy, two separate things - one is when wind farms are shut down due to local constraint problems, the other is because of national constraints e.g. when the SNSP 50% limit is breached. As far as I can discern, wind farms do not receive the type of constraint payments discussed in this blogpost, as they are not given daily schedules - instead, they get priority dispatch, if and when the wind blows (between cut in and cut out speed). However, wind energy does contribute to higher constraint payments for conventional plant as we will see.

Let's look at the annual constraint payments for the last few years :

Annual Constraint Payments paid to conventional plant
(from SEMO http://www.sem-o.com/pages/MDB_ValueOfMarket.aspx)

2010 and 2012 were years with relatively poor capacity factors for wind. The other years had average to high capacity factors. In these years, we see higher levels of constraint payments as there was more unpredictable wind interrupting the scheduled generation for conventional plant . By 2014, the payments hit a peak of € 175 million due to two factors -  saturation of wind in the system and the new interconnector which provided power from cheaper sources such as coal.  We can see which direction the payments will go when we add more wind into the mix. To put the € 175 million into perspective, SEAI claimed that wind energy resulted in € 177 million in fossil fuel savings in 2012.

EirGrid and SONI foresaw these increased constraint costs in this report :

Wind is inherently a variable resource. The UUC market schedule [done after the event], with perfect foresight, can schedule the most economic generation to balance this variability as it knows exactly the level of wind output in every period. The TSO, on the other hand, since it is not always aware of the timing or extent of these variations, must balance them using a combination of part-loaded plant and more expensive fast-start plant. This less optimal schedule will cause an increase in constraint costs. 

They then go on to say :

The generation portfolio has changed in recent years due to closures of mid-merit plant such as Poolbeg Units 1 and 2 [blogger note - these are the two plant which used the famous, now redundant, chimney stacks in Ringsend, Dublin pictured above], and the long-term forced outages of a number of other mid-merit and flexible generation units, such as North Wall 4 and Turlough Hill. In addition, transmission constraints limit output from Aghada Unit 1 and the Aghada Gas Turbines at times, further limiting the available generation portfolio. This deficit of mid-merit units that can start with relatively short notice periods has resulted in a reduction in portfolio flexibility for reacting to unexpected changes in generation and demand. Previously, when these units were available, uncertainty over generation, wind and load could be managed within 1 to 2 hours using these flexible mid-merit generator units.

At present, any potential capacity shortages due to generation, wind and load uncertainty in the future require commitment decisions to be made a number of hours in advance due to the long notice periods required by the generator units available to meet these shortages. Operators are required to call units with long notice periods further from real time when there is greater uncertainty about forecast accuracy, thus increasing the likelihood that dispatch diverges more from the optimal solution.

These commitment decisions are made to mitigate against the risk of a capacity shortage
and to ensure that sufficient replacement reserve is maintained to deal with any further
changes to generator availability or forecast demand or wind. Availability of generation with shorter notice times would mean that such commitment decisions could be made nearer to real-time and with better information. 

A provision of €3.8m has been included to account for divergence of dispatch from the
optimal solution due to generation portfolio changes. The return of Turlough Hill is expected to have a positive impact on the flexibility of the generation portfolio and the return of these generator units has been incorporated in the provision. So in otherwords, due to wind forecasting problems, there is further unpredictability in the system for conventional generators

So plant that could start up quickly (but were inefficient) were closed down and replaced with slower starting more efficient plant. This shouldn't be a problem as it should lead to more efficient use of fuel. But because it gets more difficult to forecast the wind accurately, particularly when at high penetration levels, you have to compensate these plants when they are forced off the grid by wind (which has priority dispatch) and forced to run differently to schedule.

Each plant is given a production schedule for the day. The older plants were given a schedule an hour or two before production because they could switch on quickly. But the slower modern plants have to be given their schedule up to 8 hours or more before production. The problem is that with high levels of variable wind in the system, this schedule is bound to change and plant must be compensated for these changes. Otherwise, they go out of business (which is not a good idea if you want to keep the lights on).

With wind energy now embedded in our electricity system and indeed in our energy policy, constraint payments are something we will all be paying for long into the future. It is another example of how wind adds complexity to the electricity system. A general rule in economics is that complexity costs money. You can imagine the level of office staff required to calculate the above for each and every plant in the country each day of the year. While nobody begrudges anybody a job, surely a limit needs to put on these added complexities if we want to remain competitive.

The Limit on Wind and Interconnection in the System

Graph for 18th October, 2014:

At present, the maximum amount of power from wind generation and interconnection that can be accommodated into the grid at any one time is 50%. This was confirmed by Eirgrid in their System Operational Constraints update back in March this year. In March, when the capacity factor of wind was 33%, this limit had been reached, with the excess wind power "constrained" off.  In December 2013, when wind had a capacity factor of over 50%, this limit was met on twelve different days.   Given this situation, it is curious that the current Irish energy policy revolves around more interconnection and more wind.

The problem is that wind and imported power provide "non synchronous" power which is incapable of providing the type of stable power produced from conventional sources. Exceeding this limit of 50% will alter the frequency of electricity sent to houses and factories damaging equipment and tripping the switches on the grid causing a blackout. There are efforts underway to change what is known as the "rate of change of frequency" (RoCoF) to try to accommodate more non synchronous generation but it is still not known if conventional plant can withstand these fluctuations or operate at the ranges required. Conventional generation operators claim that it could take several years before these questions can be answered. They are also worried about the costs involved with no additional benefits for them. Presumably, the consumer will eventually pick up the cost for this potentially disastrous experiment.

Eirgrid are now using a new computer tool since May called WSAT (Wind Security Assessment Tool) which monitors the amount of non-synchronous generation in the system at any one time. An example of the limit being reached recently can be seen on the 18th October (see graph above). At 14.45 during the day, wind generation was hovering around the 1,500MW mark. It had begun to rise further as demand had begun a downward trajectory. At this point demand was 2,985MW, Wind 1,561MW and the East West interconnector was exporting 54MW to the UK. The net non-synchronous power was 1,507MW, 50% of demand. Eirgrid then "shut off" 200MW of wind to bring back stability to the system. The new Great Island CCGT gas plant (presumably on testing mode) saved the day ramping up quickly to fill the gap. North Wall and Aghada OCGT also provided some additional peaking power from a more inefficient gas source.



So the next time you hear someone talking about powering the electricity grid with 100% wind or a politician talking about the need for more interconnection, you will know they are talking beyond their expertise.

Slieve Rushen Wind Farm

The Slieve Rushen wind farm, formerly owned by the Quinn Group, was recently sold for £ 100 million to Platina Partners, a London based company. This is a 54MW wind farm in Fermanagh that sells electricity to Northern Ireland in the single electricity market.

In 2008, it was revamped with 18 new Vestas V90 wind turbines and its capacity factor has been over 30% ever since. This is not surprising since it is in one of the windiest locations in the country on top of Slieve Rushen mountain.  

Records from SEM-O show that it received € 10.7 million euros in 2012 from the market in energy and capacity payments. The accounts of the company (Mantlin Ltd) are available for download in the UK companies office. These show a turnover of £ 15.6m for 2012. This converts to about € 19.3m when using the average conversion rate during the year. This means that the company receives renewable obligation credits, levy exemption certificates and curtailment and constraint payments (payments from UK grid to stop producing power) in the region of € 8.5m in total. This is quite a vast sum for activities unrelated to the generation of electricity. However, accounts show that the wind farm would have made an operating loss of £3 million as opposed to an operating profit of £3.8 million without these sources of income.

There is one more notable thing that sticks out when you read the accounts. On Note number 7, it states that the company had no employees during the year. The wind industry makes lots of claims about the number of jobs it creates, but here is one of the largest wind farms in the country employing the grand sum total of Zero employees.