City Inn Limited v Shepherd Construction Limited (30 November 2007) involved a contractor's time and money claim relating to a hotel project in Bristol. The contract in question was a Scottish JCT contract, which did not materially differ from the English JCT contract. The decision is from the Outer House of the Court of Session, roughly equivalent to the English High Court, and, we understand, will be appealed. Scottish decisions are not binding in England, but are often referred to for guidance.

This note looks at the judge's comments on:

How does one best determine where the critical path lies – by computer analysis or by a combination of evidence and common sense? 

What is the correct approach where two or more events (one being caused by the contractor, another being caused by the employer) contribute concurrently to a period of delay?

The court also dealt with a further issue relating to a "time bar" clause under which a contractor had to provide certain information as a precondition for being given an extension for time.

Retrospective delay analysis – the pros and cons of using software

The judge was faced with two programming experts who had adopted different approaches to the issue of critical delay.

The contractor's expert had originally intended to undertake a critical path analysis. However, he found it impossible to identify any critical path from the contractor's original programme. He attempted to replicate what he thought might be the logic of the original programme, but felt that this was likely to lead to unreliable conclusions. Therefore he adopted a different method. He first examined the original programme and tested it for reasonableness and completeness. He then examined the factual evidence to determine where time on the project was critically lost, and to identify the cause of that loss of time. The method was, in essence, one of "as planned against as built". He arrived at a conclusion that 11 weeks were lost, attributable to late instructions by the employer varying the works.

By contrast, the employer's expert tried to construct an as-built critical path using computer software. The judge held that this approach had serious defects. The major difficulty was that any significant error fed into the programme was likely to invalidate the entire analysis. Such errors were likely to occur if a logic link was inserted that was not wholly correct. This had happened at a number of points - as even the employer's expert had to concede at trial. That, said the judge, made his as-built critical path analysis "of doubtful value". He added: "I think it necessary to revert to the methods that were in use before computer software came to be used excessively in the programming of complex construction contracts."

The activities of installation of roof steelwork and roof coverings illustrate the two different approaches. The employer’s expert, relying on his software, said that these were not critical activities. The contractor’s expert said that they were. The judge accepted the view of the contractor’s expert. It was based on a common sense approach (it seemed unlikely that mechanical and electrical plant could be installed with the roof coverings not in place) and on factual evidence (the contractor’s records showed a sharp increase in the number of electricians and heating engineers on site in the weeks after the roof covering was put on).

This part of the judgment is a useful reminder that, while sophisticated programming techniques have their place, there are situations where they may be unreliable. In any case, delay analysis should be based on factual evidence and common sense.

The concurrent delay issue

Sometimes, more than one event can be said to "cause" a delay. As the judge recognised, it may be possible to decide between competing events by regarding one as overriding or "dominant". However, in this case, he said, none of the causes could be regarded as dominant, as each of them had a significant effect on the delay period. There were thus various concurrent causes of the delay.

What is meant by "concurrent delay" (so as to allow apportionment to take place)? In Royal Brompton Hospital NHS Trust v Hammond (No. 7) [2001], Judge Seymour QC had said that two events were not concurrent if one began before the other. The judge disagreed. Why should it matter, he said, whether one event began two days before or two days after (or at the same time as) the other? This seems to be correct. The key issue is whether the events in question caused or contributed to the same period of delay. Whether those events themselves began or ended at the same time is irrelevant. Relying on earlier case law, the judge decided that for concurrent delay he could apportion the delay between the two parties, and Clause 25 (extension of time) contemplated this. Any apportionment, (whether made by the court, or at the time of the application by the architect) had to be fair and reasonable.

In this case, there was an 11-week overrun. This was caused, the judge decided, by 11 events that were the employer's responsibility, and by two events that were the contractor's responsibility. Taking into account the significant and effect of the various events, the judge decided that a fair and reasonable approach led to a nine-week extension of time, with the contractor bearing responsibility for the other two weeks.

The apportionment approach has previously had some academic and judicial support. It will be interesting to see whether it is adopted in other cases, where no dominant cause of delay can be identified.

Time bar clauses and extensions of time

The contract featured a clause which purported to deprive the contractor of any extensions of time if it failed to provide the appropriate estimates of the effects of the variation within 10 working days of the instruction. The judge decided that, looking at this clause in context, it applied to instructions generally, but not to late instructions which, by reason of their lateness only gave rise to a need to adjust the contract sum or grant an extension of time. He accepted however that the clause had not been well drafted.

There is an argument, based on Australian case Gaymark Investments Pty Ltd v Walter Construction Group Ltd (2000), that time bar clauses like this cannot, in principle, be read as depriving a contractor of his right to an extension of time where the employer has caused the delay. This argument was not considered by the judge. Any definitive judicial comment on this argument (the judge's anti-Gaymark comments in Multiplex Constructions (UK) Ltd v Honeywell Control Systems Ltd (number 2) [2007] were obiter) will have to wait another day.