Utilities are increasingly seeking to impose additional integration charges on wind and solar generators to compensate for variability and uncertainty. Renewable Reserve Sharing Groups may provide a way to help reduce or eliminate those charges.
There are two distinct ways that wind and solar generation can impact utility balancing requirements:
- Minute-to-minute variability of wind and solar generator output adds to load variability and increases Balancing Area regulation requirements.
- Large, but rare, drops in wind or solar output, under adverse weather and load conditions and during limited availability of conventional generation, can exceed the ramping capability of online generators to follow the changes in output.
While minute-to-minute variability of wind and solar generator output tends to be minimal and is reduced through diversity as the renewable generation fleet grows, the potential for large but infrequent drops in solar and wind generation has been identified in some resource adequacy studies as a justification for addition reserves and the basis for increased variable renewable generation charges. However, an increased reserve requirement and increased costs may be based on a few minutes of insufficient ramping capacity over ten years. The studies model the power system under thousands of sets of conditions covering decades of potential weather conditions, multiple economic load growth and fuel availability forecasts, and a range of other conditions. In some cases, insufficient ramping capacity from the utility’s online conventional generators during a single five-minute interval in ten years of simulation leads to an increased reserve requirement and increased costs. Moreover, these charges are often proposed in cases where the wind and solar generation fleet is expected to grow significantly but where actual operating data at high levels of renewables does not yet exist, and the benefits of geographic diversity and aggregation in reducing ramping requirements have not yet been experienced by that utility.
While sudden drops in wind or solar generation output are similar to conventional generation contingencies in that they are rare events but potentially serious because of their size, they are unlike conventional contingencies – and do not qualify to make use of contingency reserves – because they are ramping events, taking tens of minutes to unfold, in contrast to the instantaneous tripping of large generators that create conventional contingencies. Within the traditional utility balancing framework, these potential large drops in wind and solar generation are often compensated with increased minute-to-minute regulation and ramping reserves; consequently, the calculated added reserve costs, and the added renewables integration charge, can be large.
Potential Use of Renewable Generation Reserve Sharing Groups
Renewable Generation Reserve Sharing Groups may provide a way for wind and solar generators to address high integration costs. Since generator failures are rare—it is unlikely that two generators in two different Balancing Authorities will fail at the same time—utilities reduce their conventional contingency reserve requirements by joining Contingency Reserve Sharing Groups. The North American Electric Reliability Corporation (NERC) allows Balancing Authorities to pool their contingency reserves and carry only enough reserves to cover the largest credible contingency within the combined area. This greatly reduces the amount of contingency reserves that each utility has to supply.
This reserve sharing concept could be extended to cover wind and solar reserve requirements. Recognizing that wind or solar generation is unlikely to suddenly decline in multiple balancing areas simultaneously, if only because clouds and weather fronts are finite in size and geographically can’t simultaneously cover a large area, reserves could be shared among Balancing Authorities. As with contingency reserve sharing, this would reduce the reserve requirements, and costs, for each of the participating Balancing Authorities.
Three Types of Reserve Sharing Groups Currently Recognized by NERC
At this time, NERC recognizes three kinds of Reserve Sharing Groups: Contingency Reserve Sharing Groups, Frequency Response Reserve Sharing Groups, and Regulation Reserve Sharing Groups. Contingency Reserve Sharing Groups (simply called Reserve Sharing Groups or RSGs in NERC standards) are the oldest and most common of the three types of Reserve Sharing Groups. Multiple Balancing Authorities reduce their individual contingency reserve obligations by sharing resources. This works because individual generators in the multiple Balancing Areas are unlikely to fail simultaneously. The entire group needs only to have enough contingency reserves to cover a single failure at a time. NERC lists 11 Contingency Reserve Sharing Groups.
NERC reliability standards recognize Frequency Response Sharing Groups where members pool resources to meet their frequency response obligations. The Western Frequency Response Sharing Group (WFRSG) is currently the only FRSG recognized by NERC.
NERC reliability standards also recognize Regulation Reserve Sharing Groups for the pooling of minute-to-minute regulation reserves and meeting a collective regulation requirement. Again, diversity in minute-to-minute variability among the individual Balancing Area members reduces the overall regulation requirement. There are currently no active Regulation Reserve Sharing Groups recognized by NERC.
Renewable Reserve Sharing Groups are very similar to Contingency Reserve Sharing Groups, though operating over a slightly longer time frame. There does not appear to be any technical reason that NERC would not allow RRSGs to form.
Forming Renewable Reserve Sharing Groups
NERC approval of the concept and establishment of implementation rules is first required.
Two routes are possible, using existing groups and forming new ones. Existing contingency reserve sharing groups could easily expand their services to offer Renewable Reserve Sharing. This would involve expanding the reserve resources since they could not utilize the contingency reserves; however, the same entities that supply the contingency reserves likely have additional responsive reserves they could offer to the pool.
Alternatively, wind and solar generators could form Renewable Reserve Sharing Groups, obtaining responsive reserves either through ownership or through contracts. Potential resources could include fast-start generation, on-line generation, demand response, and storage. Response from the Renewable Reserve Sharing Group would then be offered to each of the host utilities in lieu of paying the renewable integration charge.
In Summary
Contingency reserve sharing groups have worked extremely well at reducing the cost of reliably responding to large, infrequent, conventional generator failures by spreading the risk over multiple participants. A similar approach could work for wind and solar generators. If existing reserve sharing groups are unwilling or unable to expand their services to cover rare wind and solar ramping events (with additional resources), it may be possible for wind and solar generation owners to create Renewable Generation Reserve Sharing Groups as an alternative approach.
Brendan Kirby, P.E.
GridLab Consultant
For ease of sharing the information, maybe we could clarify and associate the Contingency Reserve, Frequency Response Reserve, and Regulation Reserve with primary, secondary and tertiary reserves. Then we can also examine what additional reserves the renewables need to mitigate their variability. Could they get it from those reserves or do they require an additional dispatchable product providing flexibility and ramping at the right time mitigating among others a cyclical variability ? Again this new product could be procured through the markets.
Thanks Brendan, this is a great topic. A couple of questions:
Are all of the established NERC reserve sharing groups outside of the ISO/RTO market regions? The RTOs essentially provide a reserve starting group for their members, I would think. Is that how you look at it? Is that how NERC looks at it?
Also, given the obvious benefits to ratepayers, should NERC or FERC require interconnection-wide reserve sharing groups? Isn’t there a conceptual parallel here with the BAL-003 frequency response obligation requirements that apply to all balancing areas in the interconnections?
Also, what stands in the way of allowing major wind/solar ramp events to qualify as contingency events? Sure, they are “slower” events, but why couldn’t contingency reserves be used for them (and replenished in a way that is comparable to how the faster contingency events are treated)?
Hi Mark,
Contingency reserves are one of my favorite topics. Of course, if you trip the transmission line or the unit transformer for a wind or solar plant that does count as a contingency and contingency reserves can be used. The (rational) argument for not allowing contingency reserves to be used for wind and solar ramp events concerns the event speed. A conventional generation contingency happens instantly when a unit trips (a few events are slower I a generator gets in trouble and knows it is coming off line, but still, it eventually trips even if there is a ramp down to minimum load). NERC rules give the BA 15 minutes to restore the balance from the instant the contingency began. Normal balancing requirements are suspended during the contingency. NERC rules also require the reserves to be restored within 90 minutes of rebalancing – 105 minutes total until you are back to “normal”. So it is a little tougher to say when a wind or solar event starts the 15 minute clock. I suppose the start could be when the BA operator declares that it needs help and is in an event. Even so, you would have to declare the start of the event, use the contingency reserves to restore balance within 15 minutes, and then restore those reserves and be over the event within another 90 minute. That is admittedly tougher if the ramping event extends over an hour or two.
That said, we do not treat contingency reserves very rationally. Contingency reserve costs are not allocated based on how often a generator trips, for example, or on the size of the generator. In fact, contingency reserve costs are not typically allocated to conventional generators at all. Eric Hirst and I published a couple of Oak Ridge National Lab reports on contingency reserves a while back. Allocating Costs of Ancillary Services: Contingency Reserves and Regulation (ORN/TM 2003/152 at http://www.consultkirby.com publications) was published in 2003 and followed up on an earlier 1998 report. In it we suggested allocating costs based on generator size and showed how that could be done. If you build a new 1000 MW generator and the largest generator before that was only 800 MW you just increased the contingency reserve requirement by 200 MW. The 1000 MW generator should arguably have to cover that cost. Needless to say, our proposal was not warmly received 😊.
At a minimum one could argue that wind and solar should be given financial credit because they do not contribute to contingency reserve requirements since they are not allowed to use the contingency reserves. That argument gets countered with the argument that the wind and solar unit transformers and transmission lines are still covered. The fact that unit transformers are really rare and that there must be some savings associated with dramatically decreasing the contingency reserve frequency gets ignored.
Hi Mark,
I got the responses out of order, hopefully you can make sense of them.
I agree, the ISO/RTO concept seems to me to fundamentally include reserve sharing. And some of the NERC registered RSGs are ISO/RTOs. For example, NERC lists:
• MISO-MBHydro Contingency Reserve Sharing Group (interestingly, this is one is under MRO, RF, and SERC regional compliance enforcement)
• Southwest Power Pool, Inc. (in MRO)
• ISO-NE (NPCC)
The other listed RSGs are utilities (Duke etc)
Hi Nickie
Getting the reserves types academically correct is important, and a fair amount of work has been done there, though more is always needed. Additional data from additional operating experience also helps refine the science. In addition there is also the regulatory and tariff side that renewables developers and owners have to live with. The first priority should be to get regulations and tariffs to match the science, but that takes time.
Thanks Brendan, it is a very interesting topic. It reminds me a similar concept “cloud energy storage”, which is a shared pool of grid-scale energy storage resources.
https://www.sciencedirect.com/science/article/abs/pii/S0306261916317512
Currenlty in China there is a demostration about shared energy storage by a group of renewable energy that aviod the panelty of too much variability and uncertainty:
http://finance.sina.com.cn/roll/2019-04-17/doc-ihvhiewr6505261.shtml
Hey Brendan,
You mention that NERC approval of the concept is required. But NERC doesn’t require BAAs to hold reserve specifically for wind or solar ramping conditions or forecast errors. So why would NERC need to approve a group when there is no standard to place it in? Seems like this mostly would have to do with energy trading between regions and not necessarily a NERC required concept. If regions see the economic benefit of helping each other out in this regard I can see it forming naturally through BAA to BAA agreements. Just a thought and obviously would be good to discuss the topic in NErC forums regardless. Thanks for the thoughts!
Hi Ning,
I agree there are similarities to “cloud energy storage”. In both cases the problems are more in implementation rules than with technology. The shared energy storage demonstration project is very neat. Thanks!
Hi Erik,
Interesting point. Very true that NERC does not explicitly require added wind or solar reserves. But it is concerns with NERC balancing requirements (BAL-002) that are used to justify the added wind and solar reserve charges. Specifically an imposed requirement to have the capacity to physically balance every 5 minutes – missing only one 5-minute interval in 10 years. The cost of that added capacity is imposed on the wind and solar plant. So if the wind and solar plants want to offer resources to meet the balancing need rather than simply to pay the BAA for the added reserves it seems like getting the resources judged as being technically appropriate by an 3rd party would be useful.
I agree completely that the technical solution could come from BAA to BAA cooperation. Unfortunately there may be no incentive for the host BAA to seek that cooperation – its easier just to get
the wind and solar plant to pay for added reserves – and no way for the wind or solar plants to help the cooperation happen.