Problems with using critical path analysis for proving delay claims

Date 2003

Judgment Problems with using critical path analysis for proving delay claims

In my previous article for the Legal Review, I argued that, in the case of a claim for delay on a construction project, the courts
are directing us towards the need for some sort of critical path analysis to be provided as part of the evidential material
required to prove the claim. There are plenty of planners out there that will tell you that the use of critical path analysis for this
purpose is essential. It is certainly the norm ultimately for both parties to an arbitration or litigation concerning delays to appoint
planning experts and for those experts to present their critical path model and analysis of the delays contended for. In this
article I thought that it would be interesting to examine some of the many problems associated with the requirement and use of
such techniques.

The obligations of the employer's agent when considering the contractor’s entitlement to an extension of time are usually defined
in subjective terms. For example in case of the JCT 98 Standard Form, in accordance with clause 25, if the architect is of the
opinion that the event alleged by the contractor to have caused delay is a relevant event, and that the completion of the works
is likely to be delayed thereby beyond the completion date, then the architect is required to give such extension of time for that
event as is fair and reasonable. The process of forming an opinion and deciding upon what is fair and reasonable are both
subjective requirements.

In the case of Henry Boot Construction (UK) Ltd v Malmaison Hotel (Manchester) Ltd , which considers the operation of
clause 25 under the JCT 80 standard form of contract, Judge Dyson said "It seems to me that it is a question of fact in any
given case whether a Relevant Event has caused or is likely to cause delay to the Works beyond the Completion Date . . . .
".

This seems to me to be where so many extension of time claims go wrong. The basic factual matrix is often not sufficiently
established or understood when the claim is prepared, with the result that the proper context in which an event occurred is
often ignored when alleging that an event has delayed the completion date. This is before deciding whether anything can be
added by the use of critical path analysis.

It is clear that in order to determine whether an event affected the completion date, it is necessary to determine whether the
event affected the critical path. The question is whether the architect has to make use of critical path analysis to inform his
opinion about entitlement or whether he can simply form an impression about what the critical path is at the time of the event.

In John Barker Construction Ltd v London Portman Hotel Ltd it was said that in exercising his duty under clause 25 the
architect or contract administrator must undertake a logical analysis in a methodological way of the impact of the relevant events
on the contractor’s programme. The application of an impressionistic rather than a calculated and rational assessment is not
sufficient. However, the question of what was required of the architect when operating clause 25 of JCT 1980 was also
explored in The Royal Brompton Hospital NHS Trust v Frederick Alexander Hammond and others. This was a case in which it
was alleged that the architect had been negligent in awarding extensions of time. HHJ Richard Seymour QC said that:

‘it was plain from the evidence called at the sub-trial on behalf of the Claimant, in particular that of Mr Gibson , who, of
course, is a programming expert, that there are a number of established ways in which a person who wishes to assess whether

1
a particular event has or has not affected the progress of construction work can seek to do that. Because the construction of a
modern building, other than one of the most basic type, involved the carrying out of a series of operations, some of which,
possibly, can be undertaken at the same time as some of the others, but many of which can only be carried out in a
sequence, it may well not be immediately obvious which operations impact upon which other operations. In order to make an
assessment of whether a particular occurrence has affected the ultimate completion of the work, rather than just a particular
operation, it is desirable to consider what operations, at the time the event with one is concerned happens, are critical to the
forward progress of the work as a whole. On the evidence of Mr. Gibson and Mr. Luder [for the Claimant] the establishment of
the critical path of a particular construction project can itself be a difficult task if one does not know how the contractor planned
the job. Not only that, but the critical path may well change during the course of the works, and almost certainly will do if the
progress of the works is affected by some unforeseen event. Mr. Gibson frankly accepted that the various different methods of
making an assessment of the impact of unforeseen occurrences upon the progress of construction works are likely to produce
different, results, perhaps dramatically different results. He also accepted that the accuracy of any of the methods, in common
use critically, depends upon the quality of the information upon which the assessment exercise was based. All of this does, of
course, emphasise the vital point that the duty of a professional man, generally stated, is not to be right, but to be careful. . .
. . . His conduct has to be judged having regard to the information available to him, or which ought to have been available to
him, at the time he gave his advice or made his decision or did whatever else it is that he did.

The above passage raises a number of interesting points. The first is that it is far from clear that a critical path analysis is
always required or merited. In the case of certain types of delay it may be so plainly obvious that a delaying event has
affected the critical path that provided such an event is properly evidenced and based in fact and is a relevant event then the
contractor will be entitled to the time. It may well be valid for the architect to form an impression of the critical path and the
effect of a delay on that path without undertaking his own critical path analysis. In practice of course it is preferable for the
contractor to present some form of critical path, ideally on an as-built programme, so that there is an indication of which
activities drove the completion date. Further if the decision of the architect is not accepted then a third party tribunal, who has
no prior knowledge of the project, has to be persuaded of the merits of the contractor’s claim generally through some sort of
delay analysis.

The second point is that there are various techniques for the analysis and presentation of delay claims, and it is essential to
recognise that these various methods can produce different results. I would go further and make the point that in the hands of
two different delay analysts the same method of analysis will almost certainly produce different results, because of the many
variables and subjectivity involved in such analyses.

I now turn to the commonly used techniques. The SCL ‘Delay and disruption protocol’ advocates, during the course of the
works, the regular updating of the programme and the impacting of recognised delays on the updated programme to establish
the effect of those delays at the time the events occurred and thus reflecting the as-built progress up to the time of the event.
The outcome is a reflection of the activities and logic of that part of the planned programme which at that time remains to be
completed. This is described as the time impact analysis. This method is often used for retrospective delay analysis and is
regarded by many as the preferred method because it does consider events at the time they occur and thus may give a better
measure of the contractor’s entitlement. It is however fraught with problems, not least is that as the SCL protocol warns, it is
‘the most time-consuming and costly when performed forensically’. Further there are very significant problems with communicating
the validity of the results. Two programmes are generally produced at each update, one that reflects progress before the delays
are impacted and one after. So a two-year project with monthly updates can result in some 48 iterations of the programme.
This is not very manageable.

Another significant problem with time impact analysis is that the result is heavily influenced by the quality of the base
programme, which has to be checked and verified as being achievable, and may ultimately bear no resemblance to what actually
happened. Since many construction programmes are not resourced, and there may be a variety of ways of programming the
project, this verification exercise may well be highly subjective. A further problem is that the updates of the programme require
adequate and consistent progress information at each update. If there is no such progress information or it is not reliably
consistent then the effect of a delay may be far more or far less in reality than would result from the analysis. Further key

2
factors are the way in which the events are impacted on the analysis and the extent to which any subsequent reprogramming or
resequencing is reflected in that part of the planned programme that remains to be progressed. This ongoing tinkering with the
logic at each update, which may be necessary to reflect the inadequacies in the base programme can make the steps taken by
the analyst using this technique difficult to follow even for the trained eye.
Collapsed as built requires the construction of a logic linked as-built
programme.
The protocol also talks about the collapsed-as-built or ‘but for’ technique. This method requires the construction of a logic
linked as-built programme. Delays are then added to the programme by substituting some of the as-built durations for periods of
delay. The logic is linked through and around the delays so that when the delays are ignored, the as-built programme collapses
and the as-built programme ‘but for’ the identified delays then results. I have used this technique to great effect, and it is
cheap and relatively simple to undertake. Its simplicity is also an advantage in its presentation. It has a number of drawbacks
including the difficulty in establishing retrospectively the as-built logic, the criticism that when identifying periods of delay other
work to an activity might have been going on at the same time, and that by removing the event causing delay it can be
argued that other options could have been available to the contractor, which is obviously subjective. For using impactions
on workable schedule
The planned impacted method utilises the planned programme and impacts delays on it. This requires the planned programme to
be verified and ideally to be a relatively close reflection of the actual sequence of work. It also takes no account of as-built
progress so totally ignores any contractor culpability that may be concurrent with or subsume the effect of the events alleged to
have caused delay. This in my view is the least credible technique, although it must be accepted that in the absence of
adequate as-built data, this may be the best that can be done. To use ' as-planned impacted' method if no choice.

The techniques described above are commonly used in practice, although there are available other variants of these techniques
and methods of presentation. In summary, the blind application of any of these techniques is never appropriate in my view. It is
always essential to consider the audience for the analysis, the cost benefit of adopting a particular method, and to find ways of
communicating the results in a way that is both credible and readily understandable. It is always the primary object of such
analyses to persuade in the most easily communicable way that the events complained of had the effect contended for. This
requires presentational as well as analytical skills, and a compromise between the cost of the analysis and the persuasiveness of
the results is usually required.