Pacing or concurrent delay, and what the records have to show

01 Two labels for the same slow work

Concurrent delay, as the SCL Protocol uses the term, needs two delay events, one at the owner's risk and one at the contractor's, each of which is an effective cause of delay to completion. Both have to affect the critical path (Core Principle 10). The Protocol notes that true concurrency, two events at the same time with their effects felt at the same time, is rare, and that the term is more often used for events that arise at different times but whose effects are felt together (paragraphs 10.3 and 10.4).

Pacing is different in kind. AACE describes it as a conscious, voluntary and contemporaneous decision to slow one piece of work against a delay that has already happened elsewhere on the job. That earlier delay is called the parent delay. What separates the two ideas is choice: concurrent delays happen independently of each other, while paced work is slowed deliberately by a party who knows about the other delay (section 4.2.F).

The reasoning behind pacing is practical. Once the parent delay has pushed the finish out, other work has more time than it needs. There is no point running a full crew to finish early and then wait. Slowing the work, reassigning the crew, or leaving site and coming back saves cost. In AACE's terms, the paced work is using float that the parent delay created on its path, relative to the path that carries the parent delay.

02 Not every slowdown is pacing

AACE separates two situations that both get called pacing (section 4.2.F).

• Direct pacing. Work slows because the work it depends on in the logic is running late, as when one trade can only move as fast as the trade ahead of it. AACE accepts that this is pacing in the ordinary sense, but does not treat it as a pacing delay, because the two activities are in sequence rather than concurrent.
• Indirect pacing. The paced work has no logical link to the parent delay at all. It shares the same weeks only because of timing. AACE's example is a landscaping crew that leaves site and returns later because critical work in the building is behind. This is the kind that matters when concurrency is alleged.

AACE also says that calling pacing a defence is a misnomer. Properly done, pacing is a decision taken during the work in response to the other party's delay, not an excuse found afterwards. That distinction is the whole of the evidential problem, and it is why the guidelines in section 04 below ask what was known and decided at the time.

03 The owner delay that arrives second

Not every owner delay that runs alongside a contractor delay is concurrent. The Protocol works through an example. The contractor is already five weeks late to completion because of its own event. A few weeks later the owner instructs a variation that, on its own, would have delayed completion by two weeks. One view treats both events as causes for the two weeks in which they overlap. The other says the variation did not make the job any later than it was already going to be, so the contractor's event is the only effective cause (paragraphs 10.7 to 10.9).

The Protocol recommends the second view. Concurrency arises only where the owner's event is shown to have caused critical delay to completion, which means it was on the longest path (paragraph 10.10). The Protocol adds that delay analysis is rarely precise to the day, and that the margin for imprecision should be allowed for in deciding whether delays were concurrent (paragraph 10.11).

Read with AACE's first guideline, that the parent delay must come before the pacing delay, the order of events does much of the work. A slowdown that began before the other party's delay reached the path holding completion is hard to present as pacing against it. A slowdown that began after may be pacing, if it was a choice made in response, or it may be a delay of its own.

04 What has to be shown

Where neither the law nor the contract says otherwise, AACE gives three common sense guidelines for judging whether pacing was legitimate (section 4.2.G).

1. A parent delay that came first. Pacing cannot exist on its own. There has to be another delay that is as critical as the paced work or more so, or was believed to be, and it has to come first. AACE calls this a mandatory requirement. Showing it means comparing the float on the two paths, and AACE suggests the near-critical threshold as the benchmark for when pacing needs to be analysed at all.
2. The ability to return to normal pace. The party that paced has to show it could have resumed a normal rate if needed. That assumes it could reasonably tell when the parent delay would end.
3. Evidence of intent at the time. A record that the slowdown was a deliberate decision, made when it was made. AACE accepts that such notices are rare and says this should not be a strict requirement of proof. It still recommends giving notice of an intent to pace to the party responsible for the parent delay, because without one, pacing can be put forward after the fact as an excuse for concurrent delay.

The Protocol makes the same recommendation in its guidance on mitigation: a contractor that intends to pace non-critical work should tell the owner and the contract administrator, and give its reasons (paragraph 15.2).

Contract terms come first. AACE's guidelines apply only in the absence of clear law or prevailing contract language, and AACE notes that a contract which reserves float to one party may effectively rule pacing out (section 4.2.F). The float, notice and mitigation clauses should be read before the schedule is.

05 The records that decide it

The dates in the schedule rarely settle a pacing argument on their own. The records around them do. The table sets out what tends to point each way.

6 rows, complete. These are practical indicators, not tests taken from either source.

One more check belongs with the table. AACE says concurrency is accounted for period by period, and that the overall conclusion cannot be reached by evaluating some analysis periods and not others (section 4.2.D.4). A pacing or concurrency argument built on the two or three periods that suit it should be tested against all of them.

06 Questions to put to an expert on pacing

1. Which delay was on the critical path first? Ask for the update and the data date where each one first appears on the path that held completion.
2. What float did the slowed work have when it slowed? Relative to the delayed path, in working days, from the update current at the time.
3. Were all the analysis periods evaluated? If only some were, ask why the others were left out.
4. What record made at the time shows the slowdown was a choice? A letter, an email, minutes, a revised look-ahead. Testimony given after the dispute began is weaker.
5. Could the work have returned to normal pace? Ask when the party expected the parent delay to end, and what it would have taken to restore the crew.
6. What do the contract's float and notice clauses say? They can decide the question before the schedule is opened.
7. Was the owner's delay on the longest path, or only alongside it? Under the Protocol's recommended view, an owner delay that never held completion is not concurrent.
8. Did the other party pace too? An owner can pace as well, for example by slowing its own reviews because the contractor is already late. AACE recognizes owner pacing (section 4.2.F).

07 How Critical Path Partners is set up for this

A pacing argument turns on facts the schedule holds only in part, so the work keeps the calculation and the finding apart. Each part below can be checked.

• The critical path is measured in every update, not assumed. A forensic schedule analysis records which path held completion in each period, so the order of the parent delay and the slowdown is a matter of record.
• Float is read from the updates current at the time. The comparison AACE's first guideline asks for is made in working days on each activity's own calendar, from the contemporaneous files.
• Every period is evaluated. The windows analysis runs from the first update to the last, and the periods add up to the overall result, so no period can be left out quietly.
• Concurrency is computed under each recognized approach, side by side. Which approach applies is a legal question. The analysis states what each one means for time and money and leaves the election to counsel.
• An opposing claim is tested against the standard defences. When the other side has filed a delay claim, the rebuttal analysis tests it against ten defence theories, concurrent delay and pacing among them.
• The tools leave cause blank. Whether a slowdown was pacing is a finding made by the analyst from the daily reports, correspondence and minutes, and the report names the record that supports it.

If you are deciding whether a delay claim is worth an expert at all, the free Claim Check is the first step. If the other side has called your client's slowdown concurrent delay, or has called its own slowdown pacing, the page for counsel sets out what an engagement delivers, or talk to CPP.

08 What is uncertain, and what this does not reach

• No Ontario or other Canadian decision on pacing is cited, because none has been verified for this note. How a court or arbitrator here treats pacing is a question for counsel.
• Neither source binds a tribunal. AACE says its recommended practice is not a standard of practice, and the Protocol calls itself a general statement and guide.
• The 25 April 2011 revision of 29R-03 is the one AACE's own published sample carries as of this note. Check for a later revision before relying on the section numbers.
• The Protocol's view of an owner delay that arrives second rests on English decisions, and the Protocol says it would change if an appeal court took a different view.
• The contract can change all of this. Float ownership, notice requirements and mitigation clauses come first. Under CCDC 2 the delay provisions are in GC 6.5, covered in a companion note.
• Nothing here is legal advice. The section and paragraph numbers are given so the text can be checked at source.

09 Practical points

1. Establish which delay reached the critical path first, by update and data date.
2. Measure the slowed work's float against the delayed path at the time it slowed, in working days.
3. If your client means to slow work because of the other side's delay, put it in writing at the time, with the reason.
4. Keep the ability to return to normal pace, and record when the other delay was expected to end.
5. Read the float, notice and mitigation clauses before arguing pacing either way.
6. Measure an owner delay by how much later the job finished because of it, not by how long it lasted.
7. Expect concurrency to be argued over money more than time, and prepare the separation of costs early.
8. Ask for every analysis period, not a selection.
9. Check whether the other side paced too.