The following article is Part I of a two part article on analysis of delays in a construction project. This first part explains the process of delay analysis and methodology and examines the ‘theoretical based’ methods of analysis.

The second part of the article will consider ‘actual based’ methods of delay analysis, will contrast the theoretical and actual based methods and discuss what factors may affect the selection of the appropriate method

Introduction
‘Has construction law changed much over the past decade?’ During a question and answer session at a recent lecture on construction claims, I was asked this question. In giving my answer, I considered the question from three perspectives: recent reforms, statute and case law.

With regard to the first point, I referred to Lord Woolfe’s report Access to Justice published in 1995 and his recommendations aimed at addressing the criticisms that civil justice was too slow, costly and complex. Building on the recommendations contained in this report, the Civil Procedure Rules came into effect on 26 April 1999 and replaced the Rules of the Supreme Court (the White Book). Judicial case management lies at the very heart of these reforms and I have no doubt that anyone involved in construction disputes 10 years ago will agree that the case management of construction litigation has been revolutionised over the past decade.

I then turned to Sir Michael Latham’s report Constructing the Team (more commonly referred to as the Latham Report), published in 199, a precursor to the Housing Grants, Construction and Regeneration Act 1996 (UK) (the Act) which introduced statutory adjudication into many areas of construction. Five years have passed and the effect of this legislation has been considerable. The number of construction disputes that now proceed along the traditional avenues of litigation and arbitration have been reducing each year since the introduction of the Act. Ask any construction lawyer what is the type of work they are involved in these days and the answer will probably be: ‘Another adjudication: it’s been non-stop for months’. Although it is not yet known what the long term consequences of adjudication will be, it can be stated with certainty that the introduction of adjudication has fundamentally altered the construction dispute process.

Finally, I considered recent case law, with particular emphasis on cases concerning extensions of time, which was the area of my talk. I referred to:
• Balfour Beatty Building Ltd v Chestermount Properties Ltd
• Piggott foundations Ltd v Shepherd Construction Ltd
• John Barker Construction Ltd v London Portman Hotel Ltd
• Ascon Contracting Ltd v Alfred McAlpine Construction Isle of Man Ltd
• Henry Boot Construction (UK) Ltd v Malmaison Hotel (Manchester) Ltd
• Royal Brompton Hospital v Frederick Alexander Hammond.

From these cases, the following conclusions can be drawn.
1. Theoretical calculations of delay have been discredited on numerous occasions.
2. What actually occurred on the particular project is significant when analysing contractual entitlement.
3. Who owns the program float, as between employer and con tractor, is still unclear. Simplistically one would conclude that the contractor does, but there is rarely a simple case to which to apply the rule and so the debate over ownership continues. For example, a contractor cannot allocate a general float period on a selective basis when dealing with subcontractor delays.
4. When there are two concurrent delays, one the employer’s respon sibility the other the contractor’s, the latter is entitled to an extension of time.
5. However, where the contractor is already in delay and the employer introduces an excusable event which
does not further delay the works, no extension of time is due.
6. But, if the excusable delay further delays the works, the contractor is entitled to an extension of time, on the net entitlement basis.

From the perspective of extensions of time, my view is that recent case law has not brought about significant changes to construction law. Instead, the courts have clarified a number of issues, particularly in relation to the facts and contract terms of those specific cases.

Following the lecture, I reflected on how my own practice of preparing and evaluating contractors’ claims and related expert witness work had changed over the past decade. One particular feature stood out. Trett Consulting employs three times as many forensic planners (or programmers or schedule analysts) in 2001 than it did in 1991. This is significant because it indicates how the nature of contractors’ claims has changed, in particular with respect to the analysis of project delay. There was a time when this work was the domain of quantity surveyors but today, in a firm such as Trett Consulting, half my colleagues are from a planning, construction or engineering background, rather than purely commercial. This stems from the fact that delay analysis today is more involved, more analytical, and more forensic and a more challenging feature of construction law than it was 10 years ago.

This change can also be attributed to developments in computer technology, in conjunction with more advanced planning and forensic software tools which are now available to the analysts involved in the investigation of project delays. Indeed, delay analysis has become a construction law subject in its own right, with American and UK publications available relating solely to this topic. The Society of Construction Law in the UK has drafted protocol for determining extensions of time.

I discussed how delay analysis has changed with several of my colleagues. Of particular interest were the experiences of one of our forensic planners, Chris Foan. Having given evidence in a number of arbitrations, he expressed the view that the approach to delay analysis was indeed changing. However, rather than setting out what had occurred on a project and why, he felt many claims were being presented on what should have occurred or what would have occurred but for a series of events imposed upon a party. Claims were based on theoretical models, not interrogation of fact.

We debated the number of text books now available and formed the view that while the problem has never changed (that is, ‘what caused the delay and who is responsible for the time and cost consequences?’), the means of investigating and presenting the problem (that is, the delay analysis methodologies) had changed. We decided to review a selection of these methodologies and applied them to a single project scenario. The conclusion drawn from our research was that many of the techniques produced different and, in some cases, unrealistic results which were closer to fiction than fact.

From this we coined the phrase ‘ delay analysis mythology’ and this is the theme of this article. That is to say, methodology is concerned with a set of practices, rules or procedures used by those engaged in an enquiry, and mythology is fictional story telling. Are those using delay analysis methodologies as a means for investigating project delay, actually engaged in delay analysis mythology?

The following notes briefly describe the types of methodologies in use, their advantages and disadvantages, the factors influencing selection of a methodology, the means of testing the robustness of each, and most importantly, identifying how methodology can sometimes lead to mythology.

What is delay analysis?
Before considering the various different types of delay analysis methodology, it is necessary to consider what we mean by delay analysis and why we should want to use it.

Delay analysis refers to a forensic investigation into the issue of what has caused a project to run late. That is, delay to the completion of work or contract milestones caused by the time impact of events such as variations, late information, excessively inclement weather, poor performance, remedial works and the hundreds of other delay causing circumstances that arise on construction projects.

Analysts distinguish between critical and non-critical delay, the former delaying the project’s completion date and the latter affecting progress but not overall completion. Ultimately, we need to identify those events causing critical delay for evaluating extensions of time, but it can be difficult to distinguish between each type. Analysts also investigate program disruption, or disruptive delay, which concerns issues of productivity, acceleration, congestion, fatigue, morale and other consequential effects of project change. This aspect is not considered in this article.

There are three primary reasons why one might want to analyse delay: to establish lines of investigation; to demonstrate entitlement; and to present the case one is seeking to prove.

To help establish lines of investigation
An investigation of a construction project will involve consideration of a wide variety of issues. These include: where the delays occurred (which part of the project – was it in the foundations, steelwork, roof, air conditioning and so on); when did the rate of progress decline; where did late information or materials cause delay; instances of competing delays; poor productivity; insufficiency of resources; lack of design information; failure to progress; excessive rainfall and so on.

Such an investigation requires a forensic review of project records involving three stages.
1. Databasing relevant project records, such as program activities, progress schedules, actual start and finish dates, labour allocation sheets, daywork sheets, notices of delay, material deliveries, plant deliveries, weld records, chainage progress, drawing registers and so on.
2. Analysing the databased records by linking related data together where relevant (for example, drawing numbers linked to activity progress), sorting/ grouping data by different variables (for example, progress by floor, by gang, by ‘system’, by trade and so on), aggregating or summing together quantities (for example, numbers of drawings per week, progress achieved per month and so on) and selecting or filtering data of particular interest.
3. Graphing the results of the analysis using barcharts, histograms, line graphs, tables and so on.
Analysis of project information in this way can help to highlight when events, delays or disruptions arose, how extensive they were, where they occurred on the project and which program activities were affected.

The graphs or charts produced in this way are working documents, in that their purpose is to identify changes or variances (such as peaks and troughs in resource levels, design information, overall productivity and so on), trends (indicating where delays arose, where events and delays occurred at the same time, reductions in productivity and so on) and differences (such as illustrating that certain floors/trades/systems were not affected whereas others were).

This kind of working data analysis can be raw and involves ‘slicing and dicing’ the project records in order to discover where the effects appear to exist and where the problems probably arose. In general, such analysis does not rely on a delay methodology, but requires a free format and versatile data analysis and graphics software tools. From this investigation, the analyst hopes to identify those issues, time periods or construction elements that require a more detailed study.

To demonstrate entitlement
This is the main subject of this article. I use the phrase ‘to demonstrate entitlement’ with caution as it may imply that the delay analysis using one of the methodologies is the demonstration (that is, it discharges the party’s burden to prove the consequences of a set of events upon the progress of the works). However, this implication is mythology. The delay analysis methodologies do not provide the ultimate answer in a case concerning extensions of time. The methodologies are tools for assisting in describing or analysing complex sets of facts. It is the engineer or architect, or ultimately an arbitrator or a judge, who has to consider and weigh up all the competing evidence and form an opinion. The delay analysis exercise will assist in this process but it will only be part of the evidential matrix. That is to say, the tribunal has to weigh up the terms of the contract, relevant case law, witness evidence, contemporary records such as photographs, as well as considering analytical exercises such as delay analysis, and form its own views.

To present the case
Having interrogated the project records and analysed the delays, it is necessary to convince the opposing party. Visual aids can help this process (primarily graphs and charts) and these can be produced using IT tools used by the delay analyst.

Delay analysis methodologies
There appear to be two groups, or types, of delay analysis methodology. The first category is often referred to as entitlement based methods, but this is not an ideal description since it can be confused with contractual entitlement. For example, a contractor’s entitlement to an extension of time is derived from the terms of the contract, whereas entitlement here is derived from the results of a delay analysis methodology (that is, it is methodological entitlement). The two are clearly different.

Theoretical based methods is perhaps a better definition, since these methods rely on demonstrating the theoretical impact of the consequences of delaying events, rather than on showing what in fact occurred. Another definition of this group could be model based methods, since each methodology is based on establishing a programming model of the project and then influencing it by the application of project events or constraints. The theory based grouping includes the global and net impact methods, the as-planned but for, the as-planned impacted and the as-built but for methods.

The second group is called the actual based methods, since they seek to demonstrate what actually occurred on a project and the analyst investigates what caused the project delay. These methods include the as-planned vs as-built, the window/ snapshot and the impact/update methods.

The theoretical based methods all approach the issue of entitlement to extension of time by first focusing on the delay event and then seeking to determine what delay may have resulted. However, this is not achieved by identifying its actual impact from recorded facts, but by theoretical analysis of what the effect ought to have been. These methods tend to favour the contractor’s position because matters such as culpable delay (that is, where the contractor has a problem of its own), the effects of mitigation (that is, the employer’s delay being offset by simple corrective action by the contractor) and the programming changes actually implemented by the contractor, tend to be considered only as secondary issues, if at all.

On the other hand, the actual based methods approach the analysis by seeking to measure how actual progress differed from what was planned. They focus on how the works progressed, how activities were actually delayed and only thereafter seek to ascertain what delay event(s) caused this delay.

I would emphasise that the two groups do have common features and cannot be distinguished in absolute terms. For example, the actual based methods also rely on models and theory, but less so than the entitlement or theory based methods.

One feature that all the methodologies have in common, however, is the subjectivity involved in the entire delay analysis process. If different analysts investigated the same project, applying the same method and using the same faces, they would be unlikely to arrive at the same conclusion. This is because each analyst will have to consider and challenge a wide variety of related issues and each analyst will apply different degrees of personalexperience and judgment.

These related issues include:
• the sufficiency of the planned durations;
• the sequence and logic of the planned program of work;
• the sufficiency of resources provided by the contractor to carry out the works;
• criticality, being the determination of those activities influencing progress at each stage;
• program float or risk contingency;
• the impact of revisions to program;
• concurrency of evens (two events imposed at the same time);
• concurrency of delays, (two events imposed at different times but causing delay at the same time);
• contractor mitigation, that is, making up lost time by construction experience and skill;
• dominance theory (which event was the major contributor to delay); and
• acceleration, which involves increasing the rate of production by employing more resources or working longer hours.

To take a very simple case. An element of work took two men four weeks instead of two weeks. Throughout the period the weather was dreadful. The contractor claims that the weather caused the delay. The employer asks why did the contractor not employ four men? Opinions will differ, not only between contractor and employer, but also between analysts.

The point, I believe, is this. The delay analysis methodologies each provide a set of rules for examining project delay. However, issues affecting the analysis using any methodology require subjective assessment and it is these assessments that undermine the analytical or clinical nature of the process. In addition, the rules of the
methodologies can be ill defined or require judgment in applying them,and this again increases the level of subjectivity.

In summary, none of the methodologies are perfect because they all include an element of assumption, subjective assessment and theoretical projection, some more than others. For this reason, the ‘answer’ that a methodology provides is only as good as the accuracy of the base information, the assumptions inherent in the methodology and the reasonableness of the subjective decisions made by the analyst. This is important to recognise if you are employing delay analysis to assist in presenting your client’s case.

The following notes briefly discuss each of the theoretical based methods referred to earlier. Part 2 of this article discusses actual based methods and this is followed by some practical considerations of selection and use. One problem any writer on delay analysis has is attempting to describe in words something that really requires a dynamic and graphical presentation.

Theoretical based methods

Global impact method
The global impact method is a rough and ready way of indicating what the potential impact of a delay causing event has been. An example of a global approach is where the work scope (that is, the amount of work the contractor has to carry out) doubles due to variations and so the duration of the relevant activity is doubled. In the McAlpine Humberoak v McDermott International case the claimant sought to demonstrate its delay case by arguing that every hour worked on particular variations equated to an hour delay to the project (the argument was not accepted!). Another example of a global claim arises when you indicate the date of a variation and measure the difference between the planned start or finish date of the relevant program activity with the date of the variation, to arrive at a period of delay to the project.

The global approach is quick and simple but never contractually supportable and provides no cause and effect. It ignores other delays occurring at the same time and does not consider timing, concurrency or dominance of delays. It also ignores any actual delays caused by the contractor.

This method has also been repeatedly criticised by the courts because it fails to consider the fundamental issue of criticality (that is, whether the works were delayed or not) and ignores reality, as well as the contractor’s duty to mitigate. Despite the criticisms, the global method is still widely used by contractors endeavouring to demonstrate their case.

Net impact method
This is essentially the same as the global method but with the refinement that the issue of concurrency of delays is considered. For example, where there are two concurrent delays each of five days, only five days is taken as delay to the works rather than the total of 10 days, as would be the case with the global method. The advantages and disadvantages of the global impact method also apply to this approach. In summary, this method has little to commend its use.

As-planned impacted method
This method is also known as the entitlement method and the POPE method (program of possible entitlement). It analyses the theoretical effect of impacting delay events onto the original baseline (that is, planned) program and projecting the completion date using the original sequence and timing of remaining activities. It can be used to show the theoretical delaying effect of the employer’s delays, or of the contractor’s delays, or of both together.

The prerequisites of this method are a baseline critical path program that represents the contractors intent and a schedule of delay events. The first step is an assessment of the likely critical delaying effect of each delay event in the schedule. This can be estimated using norms and experience, or be based on evidence of the actual delay experienced on the project. Second, for each delay event, the effect (such as a delayed start, delayed finish or prolonged duration) is individually impacted onto the planned program, in chronological order and the project completion date re-analysed. This process continues until all of the delays have been impacted.

The strength of this method is that the process avoids the need to analyse actual progress records in detail because the key elements of the methodology are the original baseline program and a schedule of delay events.

However, there are two principle weaknesses of this method. First, the original baseline program may not be a realistic model on which to base the whole analysis (because the works were probably carried out in a different sequence and at a different time from that originally planned). Second, since actual progress is not considered, this method does not demonstrate what actually caused delay to the works. If it can be shown that a delay event relied upon in the analysis could not have actually caused delay (for example, if it can be shown that alleged late information was received well in advance of the actual progress of the works), then the methodology will lose credibility.

If the result of this method is a projected end date that is much later than the actual achieved end date, then the reasonableness of the analysis will be in doubt, which is often the case. As with the other entitlement methods, the results derived from the analysis are likely to be attacked as artificial.

As-planned but for method
For this method, the analyst impacts the planned baseline program with the assessed implications of the events a party considers it is responsible for and the combined influences of these are analysed. The impacted completion date is then compared with the as-built completion date (that is, when the project was actually completed) and the difference is said to be how much earlier the project could have finished ‘but for’ all the other delay events (imposed by the other party) but which have not been analysed. The period between the analysed date and the actual completion date is said to represent either the contractor’s entitlement to an extension of time or the employer’s entitlement to deduct liquidated damages, depending on which set of events have been analysed.
The advantage of this method is that it is reasonably quick, as there is no need to consider actual progress of the works or the timing of events, but this means it is a theoretical investigation.

This method also relies upon the planned model for carrying out the works and ignores the ha that the actual critical path would more than likely be different. This is because a planned program is an early projection of intentions and the contractor will change the sequence and timing of activities when the works are in progress.

As-built but for method
Although this methodology adopts a model of the as-built program for analysis, and so starts off as a fact based analysis, it is still a theoretical analysis of delay. The approach is similar to the as planned impacted method, but in reverse. The as-built program is first constructed and linked together into a critical path network. This becomes the model to be analysed. A schedule of delaying events is created, including a measure of their impact (for example, a start delay, a finish delay or an activity prolongation). The very last excusable delaying event on the critical path is removed and the model is re-analysed. The difference in the overall program duration before and after this removal is said to represent the period of critical delay caused by the particular delay item removed. For example, if the last activity in the project happens to be a five day painting variation, when this excusable delay is removed, the program is able to finish five days earlier. Thus, ‘but for’ the additional painting, the program could have finished five days earlier.

This process is then continued until all excusable delays have been removed and the model has been fully collapsed (the method is also referred to as the collapsed as-built method. The process can also be applied by removing all the excusable events together, or by grouping similar excusable events together and removing each group individually. Whatever the approach, each invariably produces a different answer! Once the analysis is underway, the model no longer represents the real as-built program but is only a simulation of what the as-built program could have been had the delay events not occurred. The model is, therefore, sometimes referred to as the simulated as-built program.
The accuracy of the analysis will depend on the quality of the information on which it is based. The greater the amount of information that can be provided in support of any assumptions made, the more credible the results. Such information will generally be gained from site records, in whatever form they exist, and the importance of accuracy, completeness and reasonable logic cannot be over stressed. However, since the method is based on the as-built program, it appears to have a thread of truth about it.

Although the principles of this methodology are straightforward, its application is not. It is dependant on subjective logic links that were never set down and which were never agreed in any contemporaneous program. It is open to criticisms of bias on the part of the analyst. Also, removing delay events (and logic links) retrospectively does not reflect the actual way the works were progressed. Hence, this methodology does not appear to be an appropriate one for most standard forms of contract. In summary, the results of an as-built but for analysis are often impossible to defend because they do no relate to what actually happened on the project.

Tony Farrow, Executive Director,Trett Consulting, Manchester.

This article was originally published by theSociety of Construction Law.

Further information about the SCL and anumber of other papers are available onSCL’s website at <www.scl.org.uk>.