Insights

Deposits, Deposits, Deposits…

 Checking in on the state of the financial security deposit market — by David Riester

Background

A year ago I wrote a long, groveling article about the shortage of capital/credit available to post financial security deposits for aspiring power plants. I focused on the dearth of letters of credit (“LCs”) and possible remedies, and made a case for the continued rise of surety bonds. This article is a retrospective on how things have evolved since I posted that article, so I strongly recommend reading it before diving into this one.

The reality underpinning my position is that interconnecting to (and generally upgrading) America’s electrical grid is one of two principal bottlenecks for the energy transition (the second being the intermittent nature of solar and wind). Peeling back the onion further still, the grid/interconnection bottleneck consists of four points. 

  1. Physical: There are geographical, topographical, geological, meteorological, logistical (and more) hurdles when it comes to upgrading/evolving the electric grid. Real, tangible things need to go in real, tangible places.
  2. Human Capital (Engineering/Mechanical): Adapting/evolving the most complicated, impressive bit of infrastructure on the continent requires a huge reservoir of engineering, planning, and construction professionals. We have fewer of those professionals than are required, and suboptimal institutional structures for efficiently marshalling what we have.
  3. “Political” (Permits): Essentially all competing interests/considerations, many of which are, of course, entirely valid and worthy of consideration (land use, safety, first nations territories, public land, preservation, environmental concerns, etc.). NIMBYism falls in here too.
  4. Financial: Paying for all the upgrades, financing those costs, and assumption of liabilities.

(to conceptualize the topic at hand in the big picture…)

Supply Side Progress

Generally, the market has evolved meaningfully and positively on the supply (of capital) side. There’s been no epoch-changing innovations, but there has been encouraging incremental progress in a few areas, namely:

  1. More LC facilities (or one-offs) are closing. Not dramatically more, but more. And a few are getting done where projects/pipelines are being underwritten as opposed to a one-dimensional credit rating (far too blunt an instrument).
  2. Surety bonds are steadily expanding in availability, acceptability, and executability. The sector simply continues to grind its way to greater utility and market-share, which is as close to an “everybody wins” as one can conjure these days.
  3. Surety-backed-LCs (“SBLCs”) have finally shown up. Though I think they are superfluous byproducts of a flawed regulatory regime, in the near term, I’m glad this product graduated from “buzzword you hear at a conference and pretend to understand”, to “actual thing closing here and there”; which it has.
  4. FERC interconnection reform is helping standardize financial security deposits. So far FERC has not exactly bathed themselves in glory with FERC Order 2023’s rollout. However, this broader frustration is for another day. As it pertains to deposits specifically, the reform is nudging things towards more clarity on financial security deposits, and forcing acceptance of sureties, which is a huge win if rolled out and enforced.

The supply of solutions for satisfying financial security deposits (for IX, PPAs, swaps, forward purchase agreements) improved in 2024. The rub, of course, is that demand for these solutions increased and accelerated, rendering this specific development pathway the stickiest gear in the renewable power plant development game.

But first, the good news.

LC Facilities

As I harped on a year ago, this is the solution category with the most untapped headroom. A brief reminder of my original beef: historically, excluding those that are fully collateralized by cash deposits, LCs have been available only to large, rated parties, with few exceptions. If a bank could check the “Investment Grade Credit?” box (with respect to the strongest balance sheet guarantying the LC), then an LC is usually issued. Prior to mid-2023, there were maybe a handful of LC facilities in place for parties who could not check that box. And all LC facilities cost materially the same. In other words, it’s been excruciatingly binary for as long as I can remember.

But some rays of light have started to break through the clouds. Unfortunately, there is no aggregated data to measure the tick-up in issued LCs, so we’re left with anecdotal evidence; but those worth looking at:

Starting close to home, Segue closed an LC facility that looks to specific projects and guarantor pipeline (no IG guarantor, reasonable liquidity covenant upstairs). The bank underwrites each project/circumstance on its own merits, and has senior secured collateral for any assets where an LC is outstanding – this is the first, primary line of defense. Additionally, the Segue fund entity, which is diversified across ~200 projects, ~9 GW, and a dozen markets, guarantees the obligations – the second line of defense. 

Simple enough, but to put this in context, I’ve been attempting to secure a facility materially like the one we closed since 2015. For the first four years of that window, I was doing so on behalf of a large developer (Cypress Creek Renewables) which had a much larger balance sheet. An astute reader may rightly note that the “CCR” balance sheet wasn’t exactly oozing equity value out its pores at the time; but that didn’t stop us from raising ~$3b of tax equity and ~$750m of corporate debt/pref looking to the same balance sheet (all paid back with interest/pref for those wondering). Yet, all efforts to secure an LC facility fell flat during those years. The conversation would start promisingly, shift to shadow rating efforts (where it would stall), then shift to 100% cash collateralized “products”, and finally, after pointing out how unhelpful that product would be, we’d get kicked over to the ~12% revolver desk, which is where pipeline financing deals go to die. This cycle played out three times that I can remember. Meanwhile this recently-closed Segue facility was one of a few like it that closed over the last year. Take a look at this list:

  • Aypa Power closed a $350m letter of credit facility backed by a $100m revolver.
  • Cypress Creek Renewables closed a facility that shares features of the Segue facility.
  • Linea closed a similarly structured $145m facility; another group of weathered CCR alums who finally caught the LC unicorn.
  • Pine Gate put a few LCs in place with Zions and a broader consortium.
  • [Confidential parties] closed a $60m LC facility with an associated revolving line of credit that has commitment fees set to mimic a minimum principal outstanding amount; pipeline size is ~4 GW, balance sheet about $150m.
  • Convergent closed a multi-pronged facility with Mitsubishi.
  • Pathway Power did a deal with Rabobank focused on LCs.
  • Another Rabobank deal: Scout Clean Energy closed a $100m facility oriented toward purchase order down payments. My understanding is that this includes a combination of LCs and hard cash.
  • [Confidential parties] closed a ~$30m LC facility which has a liquidity covenant (% not known) and personal guaranty; no mention of revolver; pipeline size ~3 GW, balance sheet ~$75m.
  • Strata closed a $300m revolver/LC combo with a syndicate of banks (actually closed in late 2023 as I was ranting about LCs).
  • Vesper Energy restructured an existing LC facility into a larger one backed by a $50m revolver.

This is just the universe of publicly announced facilities plus a few extras with which I’m familiar – surely not exhaustive.

Two things jump out from this list. The first is that virtually all of these parties have a large pool of capital behind them, which can't be a coincidence. However, these deals are less reliant on those private equity shops (Blackstone, EnCap, Generate, ECP, EQT, NGP, KKR, etc.) over their shoulders than you might imagine. Maybe one with which I’m familiar features any sort of guaranty from above. That isn’t to say the presence of the robust parent is moot, of course. The presence of a big-name investor with a hefty balance sheet is a natural filter for banks – helping them validate which developers/pipelines should be taken seriously while giving them some underwriting cover. Furthermore, the availability of uncalled capital shores up the liquidity story. Yet, the upstairs balance sheets don’t feature in the documentation. Banks can’t underwrite credit exposure by simply pointing to an equity backer (parent) to which they have no recourse. These same deals were not happening even three years ago.

The second theme that jumps out partially explains the change – the market seems to have settled on a structure where a revolving credit facility sits back-to-back with the LC facilities. This is not new, per se – it’s long been a way for banks to get the LCs categorized as “collateralized” for the sake of preserving lending ratio runway, or for non-bank lenders (insurance, private credit) to facilitate LC access. But until recently this structure was primarily used to satisfy operating-phase PPA/IX securities, or debt service reserve accounts – not programmatic LC facilities. So what changed? I think banks started to figure out the back-to-back structure offered an avenue to make the margins (interest + fees) sufficiently interesting. If that’s right, I’m glad for it. My earlier article begged banks to “price in risk” such that LCs are in play for a bigger slice of the financial security needs universe. The feedback I received after publishing that piece was consistent and clear: banks, it seems, just can’t/won’t charge more than ~3% for an LC. I’ve even been told there are regulatory limits to what a bank can charge for an LC (although I’ve found no primary source evidence of this whatsoever). Whatever the underlying reasons, I remain baffled by this effective price ceiling.[1] Alas, this is the practical reality, and for many LC use cases 1–3% is simply not the right “return” for the risk or the opportunity cost of capital. The revolver component creates additional avenues to generate fees and interest, and serves to render the LC “collateralized”. Combine increased revenue with lower utilization of a finite lending runway, and it seems the product tips into the realm of the sufficiently interesting in more cases.

I’m rather pleased with this market adaptation, and trust that the demand side of the market will speak where those aggregate costs stack unpalatably. Right now, I’m not seeing that, and I encourage banks to keep heading down this path if it’s what it takes to unlock more LC capacity.

These back-to-back structures require a lot of legal paper and operational friction, but in most cases the juice is worth the squeeze if an option is available.

Whether or not the option is available for any given developer, is a complicated question. If I know developers at all, many will read the above and think “nice, so I can get a slick LC facility now, my troubles are over!” But the bar is still very high, and calls for sober realism. Would your company fit on that list from a financial and pipeline strength perspective? Remember that, in an inefficient market, there will be outliers where the stars really aligned for someone in an irreplicable way. Seeing one name on that list that you see yourself as superior to, and then spending half a year chasing a deal that isn’t there… would be a costly mistake. For those wondering if this offering is on the table for their company, the balance sheet and pipeline size of that list suggests to me that the current underwriting “threshold” is somewhere around i) ~3 GW pipelines, and ii) balance sheets above $100m. How uncalled capital is factored in severely muddies the water, so that is little more than a rough estimation of an inherently fluid, subjective line.

Surety-Backed LCs

The other roundabout path to satisfying a financial security deposit with an LC is using a “surety-backed LC” (“SBLC”). In this structure, a letter of credit is technically satisfying the financial security requirement, but the existence of the LC is contingent upon the sponsor/principal (developer) securing a payment and performance surety bond for the benefit of the LC-issuing bank.

The idea here is that there is a large universe of circumstances where i) the beneficiary (party requiring the financial security, such as a utility) does not accept surety bonds as a form of financial security, and ii) the sponsor/principal does have access to surety bonds, but does not have direct access to LCs. It’s a little easier to be deemed creditworthy enough to secure a surety bond than it is to secure an LC. So… an SBLC plugs up that gap for this category of developers.

“Surety-Backed LCs” have been a buzz word for a few years, but I’m not aware of a single SBLC actually closing before 2024. I’m sure there were a few, but it was mostly aspirational chit-chat until very recently. Sureties and banks shifted toward action in 2024.

Segue closed two SBLCs with our longest standing surety provider and a bank partner they selected. We were required to post partial cash collateral, which materially reduced the cost-of-capital benefits for Segue and our partner, and the process of securing and papering the SBLCs was a little rocky. But they closed, and all things considered these SBLCs represented the lowest cost pathway to satisfying some meaty deposit requirements. It may be the safest ~$1m of revenue I’ve seen a surety provider secure, and Segue was marginally better off with this solution vs. posting cash. Imperfect? Yes. Progress? Yes.

I’m aware of at least a handful of similar SBLCs placed this year. Additionally, intermediaries such as Twain Financial sought to productize SBLCs by gluing the different pieces together and firming up the credit story. 

SBLCs shouldn’t really be in our medium- or long-term plans. They are a rather clunky and inefficient bespoke solution. The word “stopgap” comes to mind. The two most important areas of progress are 1. Accessibility of LC facilities, and 2. Acceptance of surety bonds (by ISOs, utilities, offtakers). If those areas mature and evolve as they should, SBLCs will eventually be rendered moot. I’m rooting for that outcome, but welcome increased availability of SBLCs in the interim period, faulting no one for providing or using them.

Surety Bonds

Making surety bonds more i) accessible and ii) accepted are the lowest-hanging-fruit out there as far as relaxing the energy transition bottleneck imposed by the electrical grid and its associated interconnection queue processes. The reason is simple: surety bonds are accessible to more developers, and relatively easy to put in place from an operational friction standpoint.

FERC Orders 2023 and 2023-A require regulated ISOs and RTOs to accept surety bonds. The language is wonderfully unambiguous:

(Section 7 of 2023-A): “…acceptable forms of security for the Commercial Readiness Deposit and deposits prior to the Transitional Serial Study, Transitional Cluster Study, Cluster Restudy and the Interconnection Facilities Study should include not only cash or an irrevocable letter of credit, but also surety bonds or other forms of financial security that are reasonably acceptable to the transmission provider.”

Further clarification is offered throughout the document (an appreciative nod to Longroad Energy and Clean Energy Associates for artfully pursuing tweaks and detail).

An example (Section 185): “We are persuaded by Clean Energy Associations and Longroad Energy’s arguments on rehearing. We believe that allowing surety bonds or other forms of financial security that are reasonably acceptable to the transmission provider for the commercial readiness deposit and all study deposits will help ensure that interconnection customers do not face unjust and unreasonable or unduly discriminatory hurdles to the interconnection of new generation through limitations on the acceptable forms of financial security.”  

We’ll see if any of the ISOs, RTOs, or utilities try to get cute here – as some have in the past – by nominally allowing surety bonds but rendering them practically useless via an unusable “form of bond”, specifically the payment timing requirements. I think any such efforts will fail and prove to be a cynical waste of time. The Orders are clear, and I haven’t heard of a single case where a surety failed to pay when they should have, so it’s a losing and senseless fight to pick. 

This is a massive win for most of us. If I’d been given one wish for queue reform, I’d have wished the order forced regulated entities to accept surety bonds. The reason it’s not bigger news is simply because i) most people fall asleep at the mention of surety bonds, ii) people don’t understand financial security deposits, or what an “LC” even is; it’s consistently the biggest problem people don’t know that they have, and iii) FERC isn’t requiring everyone to turn on a dime and accept surety bonds immediately – ISOs are taking advantage of that grace period. For example, here’s what ISO-NE said in their compliance filing:

“ISO-NE has never processed surety bonds before and needs time to establish the appropriate processes and controls… the transition process, assuming FERC approval by August 12, 2024, would not allow enough time to develop the necessary processes and systems to support surety bonds.”

To be fair, I’m sure we can all agree that copying/pasting Ameren’s form bond and scheduling a 60-minute best practices call with CAISO is at least an 8-month exercise [insert eye roll emoji]. 

Demand Leaps

As alluded to, all the considerable progress above can feel hollow in the face of relentless growth in the amount of financial security required at any given point in time. Financial security burdens are increasing for well-appreciated reasons: The renewable generation and energy storage sectors are growing very, very fast; network upgrade amounts are increasing as grid saturation worsens; and ever-larger queues are leading to ever-longer queues. 

Financial Security Needed by ISO and Stage

Let’s look at the year-over-year change, the medium-term trend, and then consider whether or not it’s reasonable to expect the trend to continue, peak, or even reverse.

The graph below shows estimated financial security requirements across all US markets, applying the data above and the methodology described in Appendix 2 to previous years. The data comes primarily from Orennia[2], supplemented by Berkeley Labs’ Energy Market Policy group's 2024 “Queued Up”[3] (a truly impressive bit of work; bravo). This is a simple graph on the surface, but there are dozens of methodological nuances worth understanding in the Appendix.

In three years, the financial security posting needs imposed by ISOs/RTOs solely in connection with interconnection for wind/solar/storage only have nearly doubled. As for what portion of total financial security deposit demand this covers, I’ve not found any good data. The other need categories are i) PPAs, ii) equipment contracts (MSAs and POs), iii) swaps, and iv) financing.[4] Having had to satisfy deposit requirements for all of these, I have an anecdotal sense of frequency and magnitude. My finger-in-air estimate is that 50% of the total deposit market is interconnection related, so doubling the numbers above gets us in the ballpark. (I do not, however, think the other categories are growing at the same pace as interconnection, for what it’s worth) 

Some expected reactions to the above:

Wait, I thought the MISO queues were some of the craziest out there right now, why is that so small? Because MISO is navigating the last two queues extremely slowly. These were the queues that leaped in size (2022 and 2023 queues). Due to unprecedented study delays, both of those queues are between M2 and M3 right now, and M3 are the deposits that really start to sting. So… it’s coming. In 2025, the MISO box will be a lot bigger (I hope, otherwise they are grotesquely delayed, to the point of letting their ratepayers down even more than they are currently).

Isn’t PJM the biggest electricity market in the country? How is that not bigger? Two reasons, the first is that their interconnection process is considerably less severe on study deposit requirements (they deserve some credit for this). The second is that there have been no live queues for four years. PJM essentially shut things down to start their queue reform transition process, but they are about to fire things up again with slightly steeper readiness deposits, so this box will grow a lot in the years to come.

Will interconnection related deposits increase forever? Of course not. This reminds me of global population projections. If you look at a historical graph, you might reasonably think that global population will continue increasing, parabolically, forever; yet every single expert on the subject predicts that the population will peak somewhere between 10-12 billion near the year 2100.[5] I think our interconnection queues and their associated deposit requirements will follow a similar pattern, leveling out (maybe even shrinking) in a few years. My back-of-envelope projection:

I expect the leveling-out to arise primarily due to i) transformer, breaker, switchgear costs receding from their anomalous COVID-prompted spike, ii) queue reform, and iii) market reactions to queue saturation.

I know it asks a lot of one’s imagination to picture a near-future where the interconnection queues shrink. And maybe they won't. But the "Normalcy" or "Inertia" bias is in play here: things are hard to envision before they happen, and we overestimate the probability of an existing trend continuing indefinitely. So, I do think queue shrinkage is possible, but time and costs are the likelier drivers of deposit-needs plateauing in the near-term. It’s easy to forget we remain in the lingering throes of a secular shock to the grid component supply sector. While the demand for medium and high voltage transformers, circuit breakers, and switchgear remains intense, manufacturing throughput and supply chains are catching up to demand and recovering from the shocks from earlier this decade. We can expect a bit of cost relief in the near term as these come down to earth.

The speed at which the queues move is a critical input to total financial security need. On that front, queue reform should help. FERC Orders 2023/2023-A (the regulatory foundation of “queue reform”) raise the barrier-to-entry in the form of “readiness deposits” and “commercial readiness” tests (namely site control), while also imposing timeline (“shot clock”) requirements on ISOs/RTOs (and, by extension, utilities). So fewer projects will attempt to interconnect, and those that still do should have marginally shortened periods during which financial security deposits are outstanding. Arguably, the cluster study processes forced by FERC will reveal the infeasibility of doomed projects earlier in the queue process than has been the case where serial queue processes have been (are being) used; this should also help separate the wheat from the chaff faster, causing a cleaner IX queue.

I also expect developers to adjust their strategies and behavior as the costs of having huge pipelines slogging through long, expensive queues really start to register. For 15 years, we’ve been jabbering on about interconnection queues getting longer and more clogged up with wannabe projects (always other companies’ projects, not our own, of course). As the challenge of utility scale development stiffened accordingly, the success rate[6] of developers has suffered.

I reckon some readers might be surprised by that claim. After all, many developers have flourished during this period, and capital has flooded in. There are no aggregate metrics – at least none I can find – to definitively measure the returns on capital invested into renewable power development platforms.[7] So, though I’d unhumbly submit my career path and current perch render me as qualified an observer as any, I can’t prove my hypothesis. But here’s how I reconcile the dissonance between i) my claim that more developers are failing to be self-sustainably profitable and cashflow positive in the aggregate, and ii) the experience of watching many specific developers thrive: Humans disproportionately notice and react to other humans’ successes (as compared to their failures). This is colloquially known as “the FOMO Bias”, and academically known as the “Gluckschmerz Bias”. Humans derive more enjoyment from observing other’s failures (“schadenfreude”) but are more prone to notice other’s successes (“gluckschmerz”). Additionally, the last ~12 years’ favorable economic backdrop and increasing emphasis on socially responsible investing have provided thrust with reduced drag for those wanting to venture into the strengthening headwinds. As a result, utility scale development has accelerated, exacerbating the phenomenon of saturated queues.

But I think we’re nearing an inflection point. The primary driver is the growing pool of data/experiences demonstrating how challenging it is to create and maintain a profitable utility scale developer, and the resulting consolidation of the developer universe… which is already slowly unfurling. Over the last couple of years, a plethora of developers have “pushed on the balloon” of capital needs so hard that they have run out of money despite executing decently and assembling valuable pipelines. This has resulted in a tidal wave of developers coming to market with their hands out, needing capital for essentially the same four reasons: 1. Market expansion, 2. Team expansion, 3. Desire to carry projects further (deposits!), and 4. Desire to “become an IPP”.

Many of these are (were?) highly regarded shops who can hardly be considered startups. Yet here they are in the market desperately needing money, which might understandably be interpreted as a sign that some of these business models don’t work, at least not without frequent, large debt and equity infusions that ask a lot of the capital markets. The extreme growth in deposit requirements is a huge variable in this irreconcilable formula. 

Everything has a saturation point. One way in which that’s manifesting is ISOs cancelling or delaying their clusters/windows, because they simply cannot digest more submissions. CAISO and PJM haven’t had an open queue cluster for 3+ years; through their severe delays, MISO is headed to a similar place. Other exogenous factors in the market – such as i) deposits being out longer and ii) delayed inflows from NTP/COD events – have, I believe, pushed the market to the point where developers will have no choice but to dial back their development velocity. After many years of being very out of style, strategy/market focus will back into vogue; developer pipeline sizes will start to plateau a bit; folks will become a little more realistic about what it takes be a self-funded developer. We should welcome most of that as progress. Those behavioral shifts and trends would not be harbingers of the energy transition losing steam; not at all. We need good projects to receive the capital support necessary to become operational power plants, not a scattershot of lottery tickets purchased within unsustainable business models causing chaos at the top of the queue funnel.

Conclusion

The availability of good solutions for clean energy projects to satisfy deposit requirements is poor, but improving. The clean energy market cannot currently satisfy the amount of financial security deposits necessary to advance the energy transition at the rate required to halt climate change in time… but we are closer to that bogey than we were a year ago. Interconnection queue reform has set the stage for surety bonds to fulfill their potential. The pace of FERC 2023/2023-A implementation is unacceptably slow, so the palpability of this shift toward surety bonds remains pending, but it’s coming. Meanwhile, the letter of credit market is maturing and thawing out in the form of revolver-backed LC facilities and individual surety-backed LCs. The demand for interconnection deposits increased by about $2b year-over-year, which is probably more than the supply of solutions increased. However, the sizeable gap between supply and demand will narrow in the years to come. Queue sizes are nearing their peak. Those sizes will be increasingly uncorrelated with renewable energy deployment… in a good way. Less money will be wasted on subpar projects navigating the early stages of the IX queue before ultimately being withdrawn. With that, I expect the total amount of required financial security deposit support to level out. But it would do so at a very high level which our institutions, regulations, and financial instruments are not currently able to meet. I’m optimistic that the momentum of the last year will continue such that supply rises to meet the substantial (though hopefully plateauing) demand for viable financial security deposit solutions.

[1] If you have the answer, I’m listening.

[2] www.orennia.com; Subscription wall; we are heavy users of Orennia and find it to be best-in-class for data aggregation, analytics, and research.

[3] https://emp.lbl.gov/queues; dibs on their analysts.

[4] As an example, at Cypress we covered a few “equity-during-construction” minimums (the amount of equity a construction lender insists be contributed before funding their loan) with surety bonds. The projects were “oversourced” meaning from a bottoms-up perspective the lenders were fine lending the full EPC contract amount, but for debt:equity ratio requirements. The equity cash wasn’t needed, just the buffer of protection. So we got creative.  

[5] https://www.gapminder.org/topics/population-forecasts/ (rabbit hole of well-framed data, watch out)

[6] I measure this as the ability to achieve profitability and remain consistently cashflow positive without the perpetual need for third party capital infusions.

[7] Gross returns (ignoring overhead and corporate costs of capital) on investments in Projects, on the other hand, can be estimated. And those returns are quite good, albeit a severely lagging indicator due to the gap between capital outflows and capital inflows for projects not yet PIS.

Appendix 1: IX Queue “Funnel” Shapes

 

Appendix 2: Methodology for estimating IX deposit needs

  • Orennia produced a customized data set consisting of all projects (all technologies in all of the US interconnection queues, separated into four categories:
    • “Pre-study” means projects submitted to the queue without any official feasibility study having been performed. You might note that this category is empty for ERCOT, because you submit a project with a third-party feasibility study having been performed. So, by definition, there can’t really be a project in this category. Conversely, in MISO, anything that hasn’t received its DPP1 results land in this category. Given the delays in MISO, this category is huge as all of the '22 and '23 queues are in this bucket. The categorizing here is not perfect. For MISO, for example, “pre-study” understates the maturity most of these projects.
    • “Studies Undergoing” – at least one study has been received.
    • “IA Executed” – LGIA is fully executed.
    • “In Construction” means the actual work to build out the POI and network upgrades has commenced.
  • I organized this data by ISO and stage, and filtered for only solar, storage, and wind projects.
  • I assume that both “pre-study” and “in construction” cannot be satisfied with LCs/bonds. Cash only. This, in and of itself, would lead to an underestimation of security deposits seeking LC/bond solutions.
  • Conversely, I assume that ALL of the “studies undergoing” and “IA executed” capacity can – and is – satisfied with LCs/bonds. In and of itself, this would overestimate security deposits seeking LC/bond solutions, because much (most) of the non-refundable portions of these deposits are not appropriate risks for LCs/bonds. If the project in question still carries fatal flaw risk, non-refundable deposits are essentially equity-risk level development expenses like anything else.
  • I believe the above two methodological choices approximately balance each other out.
  • To estimate the financial security needs, the ISO-specific math is applied, which is entirely bespoke. For example, MISO is $4,000/MW in “studies undergoing”, and 20% of estimated network upgrades for “IA executed”. This isn’t precisely correct, but where more than 4 “steps” are being jammed into only 4 categories, I’ve collapsed a couple steps into one.
    • As another example, in ERCOT the analysis assumes that it costs $150k to enter the queue (split between the queue application fee and feasibility reports commissioned directly with engineering firms like EPE. This needs to be paid for with cash. The “IA executed” category leaps up because a developer needs to post the full projected interconnection costs before any real work is done by the utility).
  • Where the financial security is a function of facility costs and network upgrades, the analysis uses the historical weighted average of those costs (pulled from Orennia) for the specific ISO and technology mix
  • “Average Yrs In Queue” uses dates pulled from Orennia, only for projects currently in the queue. So this is not a full “lookback”, but rather a snapshot of the active queue. As such, it understates the time period one might reasonably expect to navigate entry to PIS.