A data center capacity change order is the most expensive category of change in construction, and not because the equipment is expensive. It is expensive because it is priced by one bidder, against a factory queue, on a project that cannot wait.
What a capacity change actually is
A capacity change is a decision to serve more critical load than the price was built for. Three more megawatts, a higher redundancy level, or a denser rack layout that needs more cooling than the design assumed.
It does not arrive as an engineering request. It arrives as a commercial one, from a customer, a leasing negotiation or an operator whose own requirement has moved.
What follows is not one change. It is a chain, because capacity moves in blocks rather than increments.
Another generator needs fuel storage, an enclosure, a pad, a transfer switch and a paralleling scheme that may itself have to change. Another switchgear lineup needs room, which may need a bigger electrical room, which is a structural and envelope change. More cooling needs plant, distribution and often more roof or yard.
Each element in that chain is a separate procurement with its own lead time, and they are interdependent, so the slowest one sets the date.
This is why a capacity change reads as one decision on a sponsor page and arrives as fourteen line items on a change order.
Why a data center capacity change order prices badly
Four reasons compound, and all four are structural rather than anybody fault.
There is one bidder. The electrical package is awarded, the subcontractor is mobilized, and the work is an extension of theirs. Whatever the contract says about competitive pricing on changes, there is no realistic second source for adding to a lineup already being built.
The factory queue does not care about your project. Major equipment is built to order against a backlog. An addition released in month six enters the queue in month six, not alongside the original order, and expediting is priced by somebody who knows why you are asking.
The markup stack applies to a large number. A change worth four million dollars carries the same percentage stack as one worth forty thousand, which makes the markup terms agreed at signature worth several hundred thousand dollars on this single change.
The schedule cost is usually larger than the construction cost. If the added equipment lands after the energization date, everything behind it moves, and on a data center the revenue attached to that date is substantial.
Put together, the same scope bought at signature and bought in month six can differ by a factor that surprises people who have only seen the equipment price.
The terms that decide how badly
Three clauses agreed before signature do most of the work here, and all three are ordinary requests.
The markup treatment. Whether the general contractor fee applies to the marked up subcontractor total or to direct cost only. On a four million dollar change that single sentence is worth roughly the cost of the entire review that found it. The mechanics are in change order markup a construction contract should fix.
Priced options for capacity steps. A price for the next block up, agreed at signature with a validity period and a release date. It costs the contractor an estimating exercise and it removes the single bidder problem entirely, because the price was set while there was still an alternative.
Equipment ordered against the higher case. For items where the incremental cost of the larger unit is small relative to the cost of adding a second one later, ordering up front is frequently cheaper even if the capacity is never used.
None of the three requires anybody to predict the future. They require only that somebody notices the capacity assumption before it is fixed, and treats it as the largest open line on the job.
A worked example
Illustrative figures. Not taken from any client project and not a quotation.
A facility priced for 9 megawatts. In month six the requirement moves to 11.
Direct cost of the addition: one generator with fuel and enclosure $1.4 million, switchgear and distribution $780,000, cooling $520,000, electrical room enlargement including structure and envelope $310,000. Direct total $3.01 million.
Applied markups under the contract as signed: subcontractor overhead and profit 14 percent, general contractor fee 5 percent on the marked up total, bond 1.1 percent, general conditions 4 percent of the change value. Marked up total $3.9 million.
Under terms agreed before signature, with the fee on direct cost only and general conditions measured rather than percentage based, the same change lands at roughly $3.55 million.
Difference from two sentences: $350,000. Separately, the generator enters a fourteen month queue and arrives eleven weeks after the target energization date, which is the larger problem and the one that cannot be fixed with a clause.
What to do when the change is unavoidable
Some capacity changes are the right commercial decision and refusing them would be worse. The question is how to buy one well.
Separate the equipment from the work. The long lead items should be released as early as possible, ideally before the full change is priced, because the queue position is worth more than the negotiating position on the installation labor.
Price the installation separately and afterwards. Once the equipment is on order, the remaining work is ordinary electrical and mechanical scope, and it can be priced against unit rates agreed at signature rather than as a lump.
Ask what is displaced rather than only what is added. A capacity increase frequently means something else in the design is now oversized or unnecessary, and those credits are rarely offered unless somebody asks.
Reprice the schedule properly. An extension request attached to a capacity change should be measured against the network, not accepted as a stated number of weeks, because the added equipment may not be on the critical path even when it feels like it is.
The fourth is where owners concede most easily and where the evidence is most available, since the schedule is a document that can be interrogated.
What to do before you sign
- Find the capacity assumption and treat it as the largest open line in the contract.
- Buy priced options for one capacity step up and one down, with validity periods.
- Fix the markup treatment so the fee applies to direct cost rather than to marked up totals.
- Agree unit rates for electrical and mechanical labor before signature.
- Require general conditions on changes to be measured rather than charged as a percentage.
- Establish the release date for every long lead item and mark the ones already past.
- Establish who decides capacity, and write the decision deadline into the schedule.
Item two is the one that removes the single bidder problem, and it is the least used tool on this list. A priced option is an ordinary commercial instrument, it costs an estimating exercise, and a contractor competing for the job will produce one.
Item seven is the one that owners can act on without anybody agreeing to anything. Most capacity decisions are late because nobody set a date, not because the information was unavailable.
The scope options for this reading sit in the review packages.
What we do
We price the capacity step before it is needed, model it under the contract markup terms as drafted, and show what the same change costs under alternative language. The output is a short comparison the owner can take into the negotiation while the terms are still open. We do not size the plant, which belongs to your engineer. The work sits in the cost and change exposure assessment.
Questions people ask
Why is a capacity change more expensive than the equipment price?
Because capacity moves in blocks and each block drags a chain behind it. A generator needs fuel, an enclosure, a transfer switch and paralleling. Switchgear may need a larger electrical room, which is structure and envelope. The equipment is the visible part of a much longer list.
Can the price of a capacity increase be fixed in advance?
Yes, as a priced option with a validity period and a release date, agreed while the contractor is still competing for the work. It costs an estimating exercise before signature and it removes the single bidder problem that makes the same change expensive later.
Should equipment be ordered for capacity that may not be used?
Sometimes, and it is arithmetic rather than judgment. Where the incremental cost of a larger unit is small against the cost of adding a second one later, ordering up front wins even if the capacity is never used. Both numbers exist before signature.
This is general information about construction contracts and is not legal advice.