Delay Liquidated Damages: Setting the Rate, the Cap and the Security

Most delay liquidated damages rates are negotiated, not derived. A number gets proposed, it gets haggled, and it lands somewhere both sides can live with — and nobody checks it against the thing it is supposed to replace, which is the money the project actually bleeds on each day it is late. A rate that does not cover daily debt service is not protection. It is a partial refund.

The previous post in this series showed that the EPC clock and the PPA clock are set by different documents and stop on different events, and that the space between them is paid by the sponsor. This one is about the machinery that is supposed to fund that space: how the daily rate should be built, how the sub-cap and the aggregate cap interact, whether LDs are the only remedy you have, whether the clause survives a challenge at all, and what instrument actually pays when the contractor has no money left.

ℹ️ Note: This is a description of how these mechanisms work in practice, not legal advice. Enforceability is jurisdiction-specific and fact-specific, and every deal's drafting differs.

What Are Delay Liquidated Damages Actually For?

Delay LDs are a pre-agreed daily sum the contractor pays for each day it misses a contractual completion milestone. They exist because proving actual delay loss is slow, expensive and uncertain — so the parties fix the number in advance. As ConsensusDocs puts it, the rate is "intended to represent the pre-estimated loss and damage that the owner will suffer" for each day of delay.

That phrase — pre-estimated loss — is the whole design brief. The rate is not a fine and it is not a negotiating chip. It is a substitute for a damages calculation you agreed not to run later. If the substitute is smaller than the loss, you have prepaid a discount on your own injury.

Delay LDs are also distinct from performance LDs, which compensate for a plant that finishes on time but underproduces. The two should be drafted separately, and — as we will come to — the question of whether they share a cap matters more than most sponsors expect.

How Should the Daily Rate Be Derived?

Build it from the bottom up. On each day the project is late it still owes debt service, still pays fixed O&M and insurance, may owe delay damages under the PPA, and earns nothing. The daily rate should cover the sum of those, not a percentage plucked from a comparable deal.

Four components, all of them knowable before signature:

• Daily debt service. Interest continues to accrue whether or not the plant generates. If the facility has converted, principal is amortising too.
• Daily PPA delay damages. If the offtake agreement carries its own delay remedy, that liability starts running on the sponsor's side and is the most direct pass-through.
• Daily fixed costs. Insurance, site security, asset management, land lease, and any standing O&M mobilisation.
• Daily lost margin. The contribution the plant would have earned, net of variable cost. Treat any merchant component as an assumption and label it as one.

Add them and you have a defensible number with a paper trail. That paper trail is not administrative tidiness — it is the evidence that keeps the clause enforceable, which we come to below.

ℹ️ Note: Market percentage ranges for daily rates circulate widely and vary enormously between sources and jurisdictions. Treat any such range as a sanity check on a number you derived, never as a substitute for deriving it.

What Caps Apply, and Do They Stack?

Three separate ceilings usually sit above the daily rate, and sponsors routinely discover only the tightest one binds. SgurrEnergy's bankable EPC guide describes the market norms as delay and performance LD caps "around 10–15% of contract price for each," a combined LD cap of "roughly 20–25%," and an overall liability cap "commonly near 100% of contract price, with consequential-loss exclusions."

Layer	What it limits	Typical market norm	Who it protects
Daily rate	Exposure per day of delay	Derived from debt service, PPA damages, fixed cost, lost margin	Owner, if derived correctly
Delay LD sub-cap	Total delay damages recoverable	~10–15% of contract price	Contractor
Combined LD cap	Delay plus performance LDs together	~20–25% of contract price	Contractor
Overall liability cap	All contractor liability	~100% of contract price	Contractor
Security instrument	What is actually collectable	Bond ~10–20%; retainage ~5–10%	Owner — this is the only layer that pays cash

The combined cap is the one that bites unexpectedly. A project that is both late and underperforming draws on delay LDs and performance LDs at the same time, and a shared ceiling means the second claim is competing with the first for the same pot. ConsensusDocs notes the alternative: "a blanket liability cap limiting the contractor's aggregate liability on the entire project — whether for performance, delay, workmanship, or any combination of issues." Simpler to administer, and considerably less protective if you were counting on separate buckets.

The practical consequence is arithmetic, not drafting. Divide the sub-cap by the daily rate and you get the number of delay days you are covered for. Beyond that day, every further day is uninsured and lands on equity.

Are Delay LDs the Only Remedy You Have?

Usually yes, and usually by default rather than by negotiation. In common-law jurisdictions the starting position is that an LD clause is the exhaustive remedy for the delay it addresses — you cannot take the LDs and then sue for the rest of your loss on the same breach.

Haynes Boone traces this through Temloc Ltd v Erril Properties Ltd [1987], where the English court held the delay LD clause was an exhaustive remedy even though the rate was set at nil — the employer recovered nothing and could not fall back on general damages. The Singapore High Court reached the same conclusion in Terrenus Energy SL2 Pte Ltd v Attika Interior +MEP Pte Ltd [2023]: an innocent party "cannot claim unliquidated damages in addition to LDs which are designed to deal with the loss that has occurred."

The narrow exception is a loss arising from a different breach that falls outside the LD clause's scope. Defective workmanship, for instance, is a separate failure with separate remedies — which is why ConsensusDocs observes that delay and performance LDs "can be cumulative and non-exclusive of other types of damages (such as the cost to repair defective work)," provided the contract is drafted to say so.

Haynes Boone's practical point is that silence is the trap: parties should state expressly whether LDs are the sole remedy for delay, "except for its termination rights, if that is the parties' intention." If you assumed you retained a general damages claim and the contract never said so, you probably did not.

Will the Clause Survive a Challenge?

A delay LD clause that a court reclassifies as a penalty is worth nothing, and the reclassification turns on whether the rate was reasoned. Bradley's analysis of City of Brookhaven v. Multiplex, LLC sets out three prongs: the injury must be genuinely difficult to estimate, the parties must have intended to compensate rather than deter, and the sum must be reasonable judged at the time of contracting.

Each prong has a practical consequence for how you build the number.

Difficulty of estimation is easy to satisfy in project finance — delay loss spans debt service, offtake penalties, tax credit timing and lost generation, which is precisely the kind of compound injury the doctrine was designed for.

Intent is where casual language does damage. Bradley notes that if a party's own representative describes the provision as a "disincentive," that testimony signals penalty intent and undermines enforceability. Call it a disincentive in a negotiation email and you have handed the contractor an argument.

Reasonableness at formation is why the derivation matters. The court's phrasing is that "the touchstone question is whether the parties employed a reasonable method" to reach a sum that "reasonably approximates the probable loss," and that arbitrary daily rates applied across different project sizes fail. Bradley's drafting guidance follows directly: avoid "blanket, cookie-cutter liquidated damages provisions," keep the rate proportionate to the work, and document in writing when LDs start to accrue.

Note the asymmetry this creates. A rate derived from your actual daily carrying cost is both more protective and more defensible than a negotiated round number. The commercial interest and the legal interest point the same way, which is rare.

ℹ️ Note: SgurrEnergy recommends "fail-safe drafting so that if an LD regime is struck down, capped general damages still apply." Without that fallback, a successful penalty challenge can leave you with the Temloc outcome — no LDs and no general damages either.

What Actually Secures the Delay LDs?

A cap describes what the contractor owes. Security describes what you can collect. On a thin project-company contractor with no balance sheet behind it, those are very different numbers, and the second one is the only one that pays debt service.

SgurrEnergy lists the standard package: a performance bond "around 10–20% of contract price," retainage of "about 5–10% (modern practice is a bank guarantee or cash retention, not both)," a parent-company guarantee "where the contractor is a subsidiary," and an advance-payment guarantee. Lender practice is consistent with this — performance security is commonly required in the range of 5% to 15% of the contract sum, with bank guarantees preferred and insurance bonds increasingly accepted.

Two features of that package matter more than the headline percentages.

Form determines speed. An on-demand bank guarantee or letter of credit pays against a compliant demand. A surety bond can require establishing default first, which is exactly the dispute you were trying to avoid by agreeing LDs in the first place. The instrument you hold decides whether you are funded during the delay or litigating after it.

The parent guarantee is the real credit. Where the contracting entity is a subsidiary, the enforceable credit is whatever stands behind it. A 15% LD cap against an entity with no assets is a 15% cap against nothing.

Rising letter-of-credit costs have pushed sponsors toward alternatives including credit insurance, parent guarantees and portfolio-level risk transfer. Those can work, but they change the payment trigger, and the payment trigger is the thing to diligence.

What Does Delay Cost on the Financing Side?

The exposure the LD rate is meant to cover is only part of what a delay actually costs, and the rest sits in the credit agreement rather than the EPC contract.

Interest accrues on drawn debt, and it compounds the problem. Interest during construction is capitalised into project cost, so a delayed project carries a larger balance at commercial operation than a punctual one. That additional debt then consumes coverage for the whole term. A delay is therefore not a one-off cost recovered by LDs — it permanently increases the amount being repaid.

The availability period can expire. A construction facility can be drawn only within a defined window. A delay long enough to run past it requires lender consent to extend, and that consent is sought at the worst possible moment: when the project is already late, the technical adviser is already concerned, and the sponsor has no leverage. The extension will be granted, and it will be priced.

Commitment fees keep running. Every month of delay is another month of fees on the undrawn portion of the facility. Small relative to the LD rate, and it accrues regardless of whose fault the delay is.

The cost-to-complete test tightens. Delay is frequently accompanied by cost — acceleration measures, extended preliminaries, standby charges. Once contingency is consumed, the certification that remaining funds are sufficient to finish fails, and drawings stop until the sponsor funds the gap.

There is a fifth item that is easy to miss and is often the largest.

The security competes with everything else. A letter of credit posted to secure the delay LDs consumes capacity on the same facility that the debt service reserve wants, that interconnection deposits want, and that PPA credit support wants. A developer running several projects can exhaust its letter of credit facility long before it exhausts its equity — at which point the choice is cash-collateralising a reserve, which carries a real cost, or not posting security at all.

That makes the form of the LD security a portfolio capital decision rather than a project-level one, and it is worth deciding across the development programme rather than deal by deal.

Two practical instructions follow.

Model delay as a package, not as an LD rate. The genuine cost of a ninety-day delay is the uncovered daily exposure the worked example below computes, plus the incremental IDC on the extended drawdown, plus the extension fees, plus whatever the acceleration cost. Netting only the first against the LD rate understates it substantially.

Check the availability period against the longstop date. If the facility's availability period expires before the contractual longstop, the financing imposes a deadline tighter than any of the four contractual clocks — and it is the one nobody lists, because it lives in a different document from the ones the project team is reading.

The test is short. Derive the required daily rate, compare it to the contracted rate, convert the sub-cap into covered days, and read off the residual exposure at your expected delay.

The inputs

Set these out as named assumptions, every one labelled:

EPC contract price          $90,000,000
Delay LD daily rate         $45,000       (as contracted)
Delay LD sub-cap            10%           of contract price
Daily debt service          $38,000
Daily PPA delay damages     $12,000
Daily fixed costs           $6,000
Daily lost margin           $9,000
Expected delay              75 days

The required rate

Required_Rate = Daily_Debt_Service + Daily_PPA_Damages
              + Daily_Fixed_Costs + Daily_Lost_Margin
              = 38,000 + 12,000 + 6,000 + 9,000
              = $65,000

Against a contracted $45,000, the coverage ratio is 45,000 / 65,000 = 69%. Every delay day is 31% uncovered before any cap is reached.

Covered days and cap exhaustion

Sub_Cap        = 90,000,000 × 10%        = $9,000,000
Covered_Days   = Sub_Cap / Daily_Rate     = 9,000,000 / 45,000 = 200 days

Two hundred days looks comfortable against a 75-day expectation — which is exactly the reassurance that hides the problem. The cap is not the binding constraint here; the rate is.

Residual exposure

LDs_Recovered   = MIN(Delay_Days × Daily_Rate, Sub_Cap)
                = MIN(75 × 45,000, 9,000,000) = $3,375,000

True_Loss       = Delay_Days × Required_Rate
                = 75 × 65,000                = $4,875,000

Residual        = True_Loss − LDs_Recovered  = $1,500,000

A project that recovers its full contractual entitlement, hits no cap, and suffers no dispute still leaves $1.5m on equity. Nobody breached anything. The rate was simply set below the carrying cost.

The sensitivity that matters

Run delay days across the columns and the daily rate down the rows, with residual exposure in the cells. Two things become visible immediately: the residual grows linearly with delay until the cap binds and then grows at the full required rate afterwards, and the rate at which residual hits zero is the number you should have negotiated.

ℹ️ Note: Model the cap as a MIN() against cumulative LDs, not against the daily amount. Capping the daily figure understates recovery early and overstates it late, and the error compounds exactly where the answer matters.

To run the full version — sculpted debt service by period, PPA damages on their own escalation, and the cap binding on a date rather than a day count — prompt Dezzmond with your EPC and PPA terms and it will build the schedule and the sensitivity grid against your own cash flows.

What Do Lenders Actually Check?

Lenders do not read the LD clause for comfort. They read it to work out how much of the construction-period risk lands on equity, because that determines whether the base case survives a delay without a cash call.

• Does the daily rate cover debt service? This is the first test and it is close to binary. A rate below daily debt service means a delay produces a funding shortfall, not merely a reduced return.
• How many delay days does the sub-cap buy? Sub-cap divided by daily rate, compared against the independent engineer's downside schedule.
• Is the cap shared with performance LDs? A combined cap means a late-and-underperforming plant draws both claims from one pot.
• What is the security, and how fast does it pay? On-demand instruments are treated very differently from sureties requiring proof of default.
• Who stands behind the contractor? The parent guarantee's credit, not the contracting entity's.
• Does a delay trip anything else? Longstop dates in the PPA, tax credit deadlines, and interconnection agreement milestones all have their own clocks.

Frequently Asked Questions

How is the delay LD daily rate supposed to be calculated?

From the owner's actual daily loss: debt service, PPA delay damages, fixed operating costs and lost margin. The legal test judges reasonableness at contract formation, so a documented derivation both protects you commercially and defends the clause if it is ever challenged.

Can I claim general damages on top of delay LDs?

Usually not for the same delay. The common-law default treats the LD clause as the exhaustive remedy — in Temloc the clause was exhaustive even at a nil rate. Damages from a genuinely different breach, such as defective work, can sit outside the clause if drafted to.

What happens if the delay LD cap runs out?

Nothing further is recoverable under that head. Each additional day of delay is carried entirely by the sponsor unless a separate remedy applies, such as a termination right at a longstop date.

Should delay and performance LDs share a cap?

A combined cap is simpler and is common market practice at roughly 20–25% of contract price. It is materially less protective than separate sub-caps when a project is both late and underperforming, because both claims compete for one ceiling.

What does delay cost beyond the LDs?

Capitalised interest on the extended drawdown, which permanently increases the balance being repaid; commitment fees on the undrawn facility; the cost of extending the availability period if it expires; and any acceleration spend, which tightens the cost-to-complete test.

Does a performance bond cover delay LDs?

It depends entirely on the instrument's wording and trigger. Some are on-demand and pay against a compliant demand; others require establishing contractor default first. Check which one you hold before assuming the cap is funded.

Closing: A Cap Is Not Coverage

The delay LD regime has four independent layers, and a sponsor can lose money at any of them without anyone doing anything wrong. The rate can sit below daily carrying cost. The sub-cap can exhaust. The combined cap can be drained by a performance claim. The security can turn out to be a bond against an entity with nothing behind it.

Only one of those is visible in the headline percentage everyone negotiates. The other three live in the derivation, the drafting and the credit — and all three are knowable before signature, which makes them a modelling exercise rather than a risk.

The first post in this series showed the gap between the EPC clock and the PPA clock. This one is the arithmetic that tells you whether the gap is funded. The next takes the date ladder itself — guaranteed dates, longstop dates and the termination rights that attach to each.

Sources: SgurrEnergy — Preparing a Bankable EPC Contract · ConsensusDocs — Liquidated Damages and Liability Caps · Haynes Boone — Are Liquidated Damages the Sole Remedy for Delay? · Bradley — Liquidated Damages Provisions: The Musts and Must Nots · McCullough Robertson — Project Finance in Renewable Energy Projects · Energetic Capital — Credit Support Options for Renewable PPAs