Interconnection Queue Interdependency
A free ride?
Imagine you’re trying to hail a cab in New York City. You see one coming down the block, but someone runs out of a nearby building, flags it down, and opens the door before you do. They tell you they’re in a rush, but when they hear you’re going in the same direction and getting off earlier, they offer you to hop in for a free ride. The situation results in one of two possible outcomes: you get where you’re going for free, or this stranger lied to you, runs out of the car at a red light, and leaves you with the bill.
This somewhat contrived scenario, as much as it sounds like a scene from a sitcom episode, is a serviceable analogy for generator interconnection queues. Renewable energy project developers grapple with similar situations as a matter of course via queue priority and the processes by which interconnection costs are shared among developers. In previous articles we explored system upgrades for interconnection studies and the math behind how these costs are shared among developers. Here, we will define the queue priority concept and explain how the relationship between upgrades and cost allocation often becomes a tangled web among projects, including those projects submitted into queues years apart from one another.
Queue priority
Much has been written lately about interconnection queues in the United States. The sizes of interconnection queues have exploded, and many ISO/RTO regions face a substantial backlog of projects. Specifically,they are actively studying projects submitted years ago into their respective queues. However, in most ISO/RTO regions, projects are not evaluated on a sequential, one-off basis. Instead, they are grouped together in what is typically referred to as a “cluster study.” This approach studies their cumulative impacts on the grid, and any upgrade costs are shared amongst the cluster. Clusters are typically defined in terms of the developers’ submission of projects into interconnection queues, and studies are performed using the relevant grid models reflecting approved transmission planning processes according to the associated time frame. ISOs/RTOs define cluster “open” and “close” dates, and a developer can submit a project into the queue to be a part of the cluster associated with that time frame. For example, all projects submitted to an ISO’s interconnection queue in 2020 may be part of a 2020 cluster, and all those in 2021 may be part of a 2021 cluster.
We now arrive at the concept of queue priority: projects submitted to earlier clusters are generally assumed to be slated to come on-line earlier than those submitted to later clusters. The earlier “prior-queued” projects have “priority” over the “later-” or “lower-queued” projects. For example, let’s consider a project submitted in 2018 and evaluated in a 2018 cluster. The project causes a grid reliability issue, an upgrade is identified to fix it, and the project is cost allocated for the upgrade. If the project – and its new upgrade – move forward to sign an interconnection agreement, then projects in the 2019 cluster (and later) could make use of that upgrade. Based on our earlier analogy, the 2018 project hopped in the cab before the 2019 project and offered to pay for the 2019 project’s ride. But what happens if the 2018 project decides to hop out?
The uncertainty of interconnection costs
When developers strategize around their current and future queue projects, they look for not only the impacts and costs associated with their projects or projects in that cluster–they also seek to understand the impacts and costs of other developers’ prior-queued projects. These projects can have a significant impact on the viability of a later-queued project. If a prior-queued project proceeds and agrees to pay for its associated upgrades, their later-queued, nearby project might be able to take advantage of those upgrades without paying anything–i.e., the stranger stays in the cab and pays for your ride. If the prior-queued project withdraws, such that they’re not paying for grid upgrades that would help the later-queued project, then a later-queued project may have to shoulder those costs (depending on the project’s impacts and distribution factors) – the stranger has jumped out of the cab, leaving you on the hook for the fare.
Of course, no one can predict the future, so developers generally have difficulty knowing which prior-queued projects and upgrades will carry forward with certainty. This is especially true as queue backlogs grow, with the number of prior-queued projects increasing. Sometimes it is obvious when a withdrawal will happen, including when the proposed upgrade costs are so high that no developer could justify them. When that occurs, it’s common to see a proposed upgrade due to some violation (or set of violations) trickle from one cluster to the next, killing projects along the way as none of them want to pay for the upgrade. This is an important point in itself: socialization of costs among developers in a cluster only goes so far. When the proposed mitigation is so expensive (e.g., in the hundreds of millions of dollars), it’s not uncommon to see a slow withdrawal of any projects cost allocated to cover it. Those withdrawals may ultimately lead to the underlying problem, and therefore its mitigation, disappearing from the study, only to return again in the next cluster’s study.
A MISO example
Let’s take a look at an actual example of how queue priority affects projects. In the DPP-2018-APR cycle (MISO’s terminology for a cluster closing in April 2018), a new 345 kV transmission line was proposed between the Hazel Creek and Scott County substations for $210 million. Due to withdrawals from that cluster, the need for the upgrade went away in DPP-2018-APR. However,the issue re-emerged in the later-queued DPP-2019-Cycle, and the following projects were cost allocated for some amount of the same $210 million upgrade in Phase 2 of their cluster study (note that that they were not cost allocated for this upgrade in Phase 1):
J1190
J1218
J1229
J1270
J1298
J1313
J1315
J1349
J1359
J1365
J1371
J1416
J1438
J1444
J1446
J1456
J1468
J1471
J1474
J1477
J1485
J1315 is a good case study. In Phase 1 of the DPP-2019 cycle, it was not cost allocated at all (i.e., $0) for the Hazel Creek to Scott County line, because at the time it was assumed the DPP-2018-APR projects were paying for it. In Phase 2, it was cost allocated over $81 million for this upgrade alone due to the prior-queued DPP-2018-APR cycle projects no longer being on the hook. J1315 later withdrew from the MISO interconnection process.
Summary
It is difficult for developers to build certainty around interconnection costs. Apart from the complexities of replicating study processes to anticipate study results before the official results are posted by the ISO/RTO, they also have to contend with gauging how decisions on current- and prior-queued projects could impact their cost-allocation. In other words, they’re getting in a cab with a stranger, and if the stranger jumps out, they need to be prepared to pay or jump out themselves.*
*No cab drivers were stiffed in the writing of this article.
Questions?
If you enjoyed this article and want to learn more about some of the things we’re working on at Pearl Street, reach out to us at hello@pearlstreettechnologies.com. Thanks for reading!