There is a lot being written electricity storage these days, and capacity markets, and the integration of variable renewable energy resources. Most of the material originates with lawyers, philosophers, sales people, consultants, politicians and other economists. I invite you to listen to a professional engineer and system operator, very carefully, for 10 minutes.
The system operator is Leonard Kula P. Eng.,Vice-President, Market & System Operations and Chief Operating Officer of Ontario's IESO. If low greenhouse gas emissions are a genuine goal, the messages are pertinent far beyond Ontario. Over the first 7 months of 2017 the IESO shows ~2.8 terawatt-hours (TWh) generated by natural gas-fueled suppliers, which is only about 3.5% of all generation. The output from gas is down over 60% from the same period in 2016, which should push the greenhouse gas intensity of IESO grid generation down to around 15 kilograms CO2e per megawatt-hour - superb for grids that aren't entirely in jurisdictions blessed with plentiful hydro-electric generation sites. From a low emissions standpoint Ontario's system operator might be considered THE system operator.
Lower gas use is one characteristic of Ontario's electricity system of late.
A less positive characteristic is rising curtailment of contracted supply - notably from wind.
Earlier this year I ended a post, Can Wind and Solar be significant contributors to a low emission electricity system, with:
...many commentators...have been skeptical about the ability to operate the grid with little of the gas generation (and previously coal) online, and ready for ramping. I believe Ontario's actual operators of the system - the ones who do it on a daily basis - have a story to tell. That story would likely include improved forecasting, revised wind turbine (and perhaps solar panel) regulation to provide a programmed reactive power element, and rationalized market bid rules forcing the curtailment of wind and solar output prior to impacting nuclear units.Kula indicates the way to make variable renewable energy systems flexible is by curtailing their output.
It is variable renewable energy systems that need to be flexible.
We added the ability to dispatch the variable generation facilities that are connected to the grid in 2013, and they are very fast acting resources. At times we have to reduce the output of these facilities to go ahead and manage congestions and/or surplus conditions, and once they dispatched down they are very quick and flexible to be dispatched back up.The IESO's Kula spoke at the 2017 Stakeholder Summit. The IESO requires you register before viewing the recordings (it's a painless process). The section I am addressing is the panel on "Building Flexibility into the Electricity System". While I highly recommend the 10 minutes at the start of the panel in which Kula delivers his presentation , Rob Coulbeck is also worth listening to, as is the Q & A session.
I think listening will position people to far better understand some power plant decisions being made throughout the world, including:
- a 100 megawatt battery, with 129 megawatt-hours of energy, to be built by Tesla and paired with Neoen's Hornsdale industrial wind facility (which I understand has a 100 megawatt (MW) first stage operating and is on its way to a total 309 MW capacity).
- a 100 megawatt "battery energy storage system" being built by AES as part of a 640 MW California Natural gas-fired power plant
- the choice of twelve 18-cylinder Wärtsilä 50SG engines operating on natural gas for the 225 MW Denton (Texas) Energy Center. The project now has some controversy attached to it, but I don't think it's disputed engines were chosen over turbines due to a superior "ability of the plant to quickly start and stop." The engines are perceived as better complements to variable renewable generators.
These all seems to provide flexibility - in the case of the Tesla/Neoen batteries this is a more important aspect than storage.
On the a non-cutting edge technology point, people will also better understand why jurisdictions like Germany persist in burning coal while increasing variable renewable generators. Kula's explaination of how coal can be, from a flexibility perspective, superior to traditional natural gas generators is welcome, but hopefully this isn't a new message for readers of my work. Donald Jones (writer of the first guest post on the blog) wrote this back in 2010:
The coal plants have served Ontario well providing baseload, intermediate load, peaking load, operating reserve, and high ramping capability for load following as well as frequency control and voltage control. Coal is much more flexible than gas. It has a much lower minimum load and a shorter minimum run time than the gas-fired combined cycle gas turbine units with their large complex heat recovery steam generators. Coal has over twice the dispatch flexibility range, between minimum loading and full power, which means either twice as many gas units would be needed to provide the same dispatch flexibility range as similar sized coal units or the gas units would have to have twice the capacity of the coal units... Dispatching combined cycle plants is much more complex than dispatching coal-fired plants since the steam turbine is a slave to the combustion turbines and various combinations of the multiple combustion turbines with and without the steam turbine are possible, with warm-up of the steam generators and steam turbine to be consideredI suggest jurisdictions really learning from Ontario's reductions in emissions will hesitate to enter into long-term contracts with the types of natural gas fueled power plants Ontario did, but alternatives remain uncertain.
Appendix: Transcription of Leonard Kula presentation
Some Recent trends:Generally as a system operator we have noted a decrease in system flexibility over the last decade or so.
Hydroelectric facilities are not as flexible as they used to be, for a variety of reasons:
In Ontario we’ve also noted a change from coal to gas. Coal units were characterized by very low minimum loading points which gave them a lot of room to maneuver. Natural gas facilities have higher minimum loading points and in Ontario when you’ve got increased amounts of baseload facilities over the past few years, and decreases in demand for a variety of reasons, we actually see that we are getting our coal fleet on the line – sorry, our natural gas fleet on the line – less than we used to get the coal facilities on the line.
The picture is not all grim.
We’ve added some flexibility back into the system.
We added the ability to dispatch the variable generation facilities that are connected to the grid in 2013, and they are very fast acting resources. At times we have to reduce the output of these facilities to go ahead and manage congestions and/or surplus conditions, and once they dispatched down they are very quick and flexible to be dispatched back up.
In 2016 we completed an operability assessment at the IESO that assessed the impact of increased quantities of variable generation. We found that the accuracy of the variable generation forecast is very good within an hour, and the accuracy the accuracy of the variable generation forecast is good, but not great, one hour out when you are scheduling imports and exports in and out of Ontario; five hours out, which is the typical time when we are committing our gas units; and 24 hours out when we schedule and commit resources day-ahead.
So on this graph … the forecast error is depicted here. The sharp and steep graph is the forecast error that we see within the hour. The point of this slide is that within the hour we’ve got much better information which then feeds into what resources we have to dispatch on to the system.
- Public safety
- Stewardship concerns
- Recreation
- Erosion
- Fish spawning, and a number of other things
- Mechanical – you’ve got old equipment that may not be as able to adjust their output as frequently as it did when the equipment was newer,
- and as a system operator we frequently encounter what we call hydro-electric lock-outs: a dispatch of a change in output for a hydro-electric facility and then being unable to move it again for a number of hours.
In Ontario we’ve also noted a change from coal to gas. Coal units were characterized by very low minimum loading points which gave them a lot of room to maneuver. Natural gas facilities have higher minimum loading points and in Ontario when you’ve got increased amounts of baseload facilities over the past few years, and decreases in demand for a variety of reasons, we actually see that we are getting our coal fleet on the line – sorry, our natural gas fleet on the line – less than we used to get the coal facilities on the line.
The picture is not all grim.
We’ve added some flexibility back into the system.
We added the ability to dispatch the variable generation facilities that are connected to the grid in 2013, and they are very fast acting resources. At times we have to reduce the output of these facilities to go ahead and manage congestions and/or surplus conditions, and once they dispatched down they are very quick and flexible to be dispatched back up.
In 2016 we completed an operability assessment at the IESO that assessed the impact of increased quantities of variable generation. We found that the accuracy of the variable generation forecast is very good within an hour, and the accuracy the accuracy of the variable generation forecast is good, but not great, one hour out when you are scheduling imports and exports in and out of Ontario; five hours out, which is the typical time when we are committing our gas units; and 24 hours out when we schedule and commit resources day-ahead.
So on this graph … the forecast error is depicted here. The sharp and steep graph is the forecast error that we see within the hour. The point of this slide is that within the hour we’ve got much better information which then feeds into what resources we have to dispatch on to the system.
… Bottom line is that we get much better information to go ahead and assess the resource fleet and what we need to do to balance the system within an hour.
Given that we get much better information within the hour, we find however that there are limited available Ontario resources that are able to respond to supply shortfalls if wind and solar production is less than we expect or if electricity demand is greater than we expect.
…[quotes from Richard Carlson]
“The growth in generation from renewable generation resources such as wind and solar, creates a need for greater flexibility in the generation mix to ensure the reliability of the system.”As a consequence of our 2016 Operability Study, we assess that we need an additional – about – 840 megawatts of flexible resources in Ontario by the end of 2018. Again coming back to my original statement of what does flexibility mean in Ontario,
Flexibility is supply resources able to respond to IESO dispatch signals to increase their output within ~30 minutes.”
This shortfall of flexible resources gets reflected in market results… [discusses price spikes in 1st quarter of 2017]
…
As a system operator, we need flexible resources today. When conditions exist for short term supply:demand imbalances and we expect there could be a reliability impact, as a system operator we are bringing on resources with longer start-up times that can then provide flexibility should there be an imbalance in real-time where you need the flexibility to respond.
In the short term …[discusses these points from presentation]
• To meet flexibility needs in the short-term, we need to:
- Incent greater flexibility from existing assets and participants
- Signal the need for and dispatch flexibility
- –More frequent intertie scheduling (15 minutes)
- –Assessment of the need for a flexibility product/mechanism
• Bring in lessons-learned from other jurisdictions
spreadsheet
No comments:
Post a Comment