Concurrent delay, and why both parties think they are right

Concurrent delay is what happens when the owner causes a delay and the contractor causes a delay and both are running at the same time. It is the most argued about subject in construction scheduling, and the reason is that the answer depends on a methodological choice nobody agreed in advance.

What the argument is actually about

Suppose a project is delayed by twenty days. During the same twenty days, the owner was late releasing an area and the contractor was short of labor. Both are real. Either one alone would have caused the delay.

Three questions follow and they have different answers.

Does the contractor get an extension of time. Usually yes, in most American jurisdictions, because the owner caused delay was sufficient on its own.

Does the contractor get money for those twenty days. Usually no, because the contractor would have been delayed anyway and has therefore suffered no loss attributable to the owner.

Does the owner get liquidated damages. Usually no, because the owner cannot claim damages for a period it partly caused.

That outcome, time but no money to the contractor and no damages to the owner, is the general position and it is a reasonable one. The fight is almost never about the principle. It is about whether the delays were genuinely concurrent, and that is a question of analysis rather than of law.

Why the method decides the answer

Several recognized methods exist for analyzing delay and they produce different results on the same facts. That is not a scandal, it is a consequence of the question being genuinely hard.

As planned versus as built. Compare the original schedule to what happened. Simple, intuitive, and weak, because it does not account for changes to the plan along the way.

Impacted as planned. Insert the delay events into the baseline and see what the completion date becomes. Favors whoever is inserting, and ignores everything else that happened.

Collapsed as built. Take the actual schedule and remove the delay events to see what would have happened without them. Requires a reliable as built record, which frequently does not exist.

Time slice or windows analysis. Break the project into periods and analyze the critical path in each. The most widely respected method and the most expensive, because it requires every monthly update to exist and be reliable.

The windows method is also the only one that can properly identify concurrency, because concurrency is a question about what was critical during a specific window, and the other methods do not look at windows.

What makes concurrent delay real rather than asserted

Two delays are only concurrent if both were on the critical path at the same time. That is the whole test and it is where most assertions fail.

A contractor short of labor on an activity with eighteen days of float is not causing delay at all. Raising it as a concurrent cause is raising something that was not critical.

Equally, an owner late on a decision affecting work that was not yet on the critical path has not caused delay either, however uncomfortable the correspondence looks.

So the analysis reduces to a sequence of narrow questions. During this window, what was critical. Was the owner event on it. Was the contractor event on it. Were both.

Answering that requires the monthly schedule updates to exist as a series. Which is why the record keeping described in reading a schedule update is not administrative housekeeping. It is the evidence base for the single most expensive argument on the project.

An owner without that series is arguing about concurrency from memory against a contractor who kept every file.

A worked example

A 240,000 square foot office building completes 64 days late. The contractor claims all 64 as excusable and compensable, at $290,000 of extended general conditions plus $410,000 of associated cost.

The owner position is that the contractor was late on the structure for its own reasons.

A windows analysis over eight periods produces a different answer from either. In windows one through three the structure was critical and the contractor was behind, 31 days, contractor risk. In windows four and five a late owner decision on the facade was critical, 18 days, owner risk and compensable. In windows six and seven both a contractor labor shortage and a late owner selection were on the critical path simultaneously, 15 days, genuinely concurrent.

The outcome is 33 days of extension, of which 18 carry cost and 15 do not, and 31 days at contractor risk against which liquidated damages apply.

Neither opening position survived. Both parties had asserted a single characterization of a period that contained three different ones.

What to agree before any of this happens

Two provisions in the contract remove most of the expense, and both are agreed in minutes at signature.

Name the analysis method. A contract specifying a windows analysis using the contemporaneous monthly updates removes the entire argument about methodology, which is frequently the most expensive part of a delay dispute.

State the concurrency rule. Whether concurrent delay gives time without money, whether it is apportioned, or whether some other treatment applies. Apportionment is disfavored in many American courts but parties can agree it, and an agreed rule beats an argued one.

Alongside those, require the monthly updates to be issued in native format and retained, because a windows analysis is impossible without them and the requirement costs nothing.

Owners who do these three things convert a six figure forensic exercise into a two week reconciliation, and the saving accrues to both parties, which is why contractors usually agree to them.

The practical position during a project

Most concurrency arguments are avoidable, and the avoidance is not clever drafting. It is answering extension requests on time.

An owner who grants the 18 days that are genuinely owner caused, at the time, in writing, has removed those days from the dispute permanently. What remains is a narrower argument about the rest.

An owner who refuses everything and insists on the original date has created the conditions for a construction acceleration claim on top of the delay argument, as described in directed and constructive acceleration.

The discipline is therefore to assess each request on the window it affects rather than on the overall position. That is harder than it sounds, because the instinct is to answer the whole claim, and the whole claim is where concurrency hides.

Window by window, most delay disputes are a series of small, answerable questions. Taken as a whole, they are one large unanswerable one.

The same logic applies to the owner’s own conduct. An owner who knows that a decision is going to be late should say so, in writing, with a date attached, rather than letting the deadline pass silently. A notified late decision is a bounded event that can be planned around. An unnotified one becomes evidence of indifference in a claim eighteen months later.

None of that requires admitting fault. It requires writing down what is going to happen, which is the same habit that produces the record in the first place.

What we do

Before signature we name the analysis method, settle the concurrency rule and require the update series to be retained, which together remove most of the cost of a future argument. During a project we read the updates window by window so that the owner position on each period is formed while the evidence is current. The pre signature work is the schedule and procurement risk review. With three weeks or more before signature, the full pre-GMP review reads the price, the schedule, the interfaces and the change exposure together.

Questions people ask