Reasonability in constructing fragnets: A crucial perspective required for the resolution of any type of dispute

Reasonability¹ is a crucial perspective required for the resolution of any type of dispute. Whilst construction disputes are not an exception, practitioners in the construction industry often fail to develop a common understanding of reasonability in delay analysis. In many instances, this turns the dispute resolution procedure into a blame game. Therefore, a clear understanding and application of reasonability is a crucial factor for a fair and reliable delay analysis.

Various delay analysis methods are available to the construction industry and all of them comprise various crucial steps for their correct implementation. Many of these steps require reasonability checks during the application of the selected methodology.

Among these delay analysis methodologies, the industry and the courts recognise Time Impact Analysis carried out in windows of time (“TIA”) as one of the most robust and reliable delay analysis methodologies. If it is performed prospectively, TIA provides a structured approach for the contemporaneous analysis of delay events.

Conversely, when performed retrospectively, especially in cases where the project is complete and recognising that court’s preference is generally analysis based on the factual as-built data, TIA still supports the delay analysis process with its capability to demonstrate critical path movements, especially in complex projects, and adaptability for concurrency analysis. Where the analysis uses contemporaneous progress information through to completion it provides an encapsulation of the as-built data progressively through to completion of all of the works. Thereby fully satisfying the court’s natural preference to demonstrate what factually occurred on the project whilst still addressing those issues otherwise missed by a static analysis of the planned and final as-built positions on a project.

In relation to TIA’s capability to capture critical path movements, in Mirant Asia-Pacific Construction (Hong Kong) Limited – and – Ove Arup and Partners International Limited² Judge Toulmin stated that:

Windows analysis, reviewing the course of a Project month by month, provides an excellent form of analysis to inform those controlling the Project what action they need to take to prevent delay to the Project. 

Without such analysis those controlling the Project may think they know what activities are on the critical path but it may well appear after a critical path analysis that they were mistaken.

Similarly, in Walter Lilly & Company Limited v Giles Patrick Cyril Mackay, DMW Developments Limited³ Mr Justice Akenhead provided that:

Very few programmes were formally issued by WLC after 16 February 2007. In particular, there is no programme of all the works outstanding at that date which could sensibly be used as a baseline in a retrospective programme analysis.

It is therefore not possible to carry out a “traditional” delay analysis, which uses the Claimant’s programmes to identify the critical path during the period after 16 February 2007 in the way one might normally expect. It will instead be necessary for the experts to form a view as to what were the critical (or driving) delays in the period after 16 February 2007 without the assistance, which would normally be available from contemporaneously produced programmes.

In other words, TIA’s capabilities to demonstrate critical path movements, especially in complex projects, and adaptability for concurrency analysis make it an extremely useful tool for the assessment of delays both prospectively during the course of the projects or retrospectively upon completion.

Implementation of TIA in windows of time, the delay analyst is simply required to:

Step 1: Divide the analysis period into windows of time
Step 2: Identify the delay events and associated facts
Step 3: Model the delay events as fragnets4
Step 4: Incorporate the fragnets into the schedules in chronological order at the start of each window
Step 5: Update the impacted schedules for actual progress at the end of each window; and
Step 6: Ascertain the impact of the delay event in each window.

There are several factors that affect the successful application of these steps, and fragnet construction is one of them.

A fragnet consists of a group of activities that represent a mathematical model of a delay event. Therefore, Step 3 of the foregoing procedure unveils one of the most important questions within the delay analysis practice:

How should, the delay analyst model the delay event fragnet so that it, and its imposed impact, represent the reasonable effect of the delay event?

The importance of this query stems from the fact that fragnet modelling and insertion into the schedule largely defines the threshold where delay analysis results fluctuate from ‘hypothetical and/or exaggerated’ to ‘reasonable, reliable, and robust’.

In order to avoid unreasonable and unreliable TIA results, fragnets used in a delay analysis should be tested against a reasonability criteria that clarifies whether the fragnet should extend beyond a window cut-off date or otherwise be split into multiple ‘subfragnets’ which extend up to, but not beyond, the cutoff date of the window.

The foregoing reasonability test becomes more crucial if the delay analysis is being performed retrospectively when all the facts and the as-built information related to the delay event is available to the delay analyst.

Accordingly, this article seeks to address the above query by examining the following two delay event scenarios that are analysed retrospectively:

Scenario I: Instruction to hold design; and
Scenario II: Change in site access strategy.

Scenario I: Instruction to hold design

In this scenario, the Employer instructed the Contractor to hold part of the design on 5 January 2016.

Following design discussions that took three months, on 5 April 2016, the Employer released the hold and instructed the Contractor to proceed with the design.

The analysis periods and the actual start and finish dates of the delay event mentioned in the foregoing scenario are illustrated in the figure below: Figure 1 shows that the actual event starts in Window No. 2, extends through several windows and ends in Window No. 5.

Figure 1 shows that the actual event starts in Window No. 2, extends through several windows and ends in Window No. 5.

Accordingly, the analyst has two options to analyse the effect of this delay event, namely:

  • Impacting the Dec-15 schedule at the beginning of Window No. 2 with the fragnet constructed using the as-built dates; or
  • Dividing the as-built fragnet into four segments based on cut-off dates of the windows and impact the corresponding schedules at the beginning of each window.

The first option is not a reasonable option because, at the time of instruction, the Contractor could not reasonably foresee that the Employer would release the design hold on 5 April 2016 unless in its instruction the Employer provided the planned release dates.

Furthermore, the use of Dec-15 schedule as the baseline schedule and inserting the full as-built fragnet into it may result in hypothetical and exaggerated results similar to the forecasts estimated by an Impacted as Planned (“IAP”) methodology, which lacks credibility and court support.

In conclusion, for this scenario, the reasonable approach is the insertion of the fragnet partially in each window. In other words, in Window No. 2 the analyst should insert a fragnet that ends at the cut-off date of Window No. 2. Then, in the next window, the schedule should be updated to the beginning of Window No. 3 and the fragnet that already constructed in Window No. 2 should be extended up to the cut-off date of Window No.3. This procedure should be repeated until Window No. 5. In Window No. 5, the analyst is required to extend the fragnet up to 5 April 2016, the date when the hold was released, because this information was available to the Contractor during the period of Window No. 5.

It is crucial to note that if the Contractor was supposed to perform additional engineering activities subsequent to the hold release, the fragnet used in each window should consist of those additional engineering activities together with the subsequent review and approval cycles.

Scenario II: Change in site access strategy

In this scenario, the Employer issues an instruction to change the site access strategy.

Initially, the Contractor was supposed to obtain access to the full site at the commencement of the project on 10 December 2015.

On 15 January 2016, the Employer divided the site into zones and issued individual access dates for each zone as follow:

  • Zone A – 15 February 2016;
  • Zone B – 20 March 2016; and
  • Zone C – 14 April 2016.

As illustrated in Figure 2, the revised access dates fall into different windows.

Even though access to each zone is planned to be released in different windows, during the period of Window No. 2 the Contractor could reasonably foresee the access dates based on the Employer’s instruction dated 15 January 2016. Therefore, in order to determine the delay effect of this event, the analyst can impact the Jan-16 schedule with all instructed access dates notwithstanding that these dates extend beyond the end of Window No. 2.

A similar approach can be followed for the construction of the fragnets that model scope changes. In other words, the analyst should create fragnets based on the information available within the window under consideration. MCAA’s Bulletin No. CO35 also supports the approach explained above.

Fragnet construction perspective of MCAA can be briefly summarised as follows:

  1. Identification of the fragnet activities related with the delay event and development of the activity network based on the facts ‘that are known at the time the potential impact is identified, or can be reasonable predicted as a result of the impact event.
  2. Demonstration of each period after the trigger (start) of the delay event with a related activity in order to ‘ensure that there are no unidentified gaps in time within the fragnet’.
  3. Verification of the actual start and finish dates of the fragnet activities by investigation of the project records.
  4. Identification of the base scope activities that are affected by the delay events.
  5. Evolution of the fragnet (insertion of new activities) in each succeeding window based on the available information while actualizing the completed activities.

In conclusion, fragnets are not static modelling objects and should evolve during the course of delay analysis based on the available information and facts associated with the delay events in each window.

Factual substantiation

In addition to the use of the information available in each window, the delay analyst should also provide factual evidence that substantiates the fragnet activities and their evolution in time. This is emphasised by AACE International Recommended Practice No. 29R-03 Forensic Schedule Analysis (AACE Recommended Practice) as follows:

No forensic schedule analysis method is exact. The level of accuracy of the answers produced by each method is a function of the quality of the data used therein, the accuracy of the assumptions, and the subjective judgments made by the forensic schedule analyst …

Any analysis method, no matter how reliable and meticulously implemented, can fail if the input data is flawed…

The best accuracy that an analyst can hope to achieve is in the faithful reflection of the facts as represented in contemporaneous project documents, data, and witness statements.6 

Therefore, the analyst should ensure that the input data is not flawed and does not contradict the actual history, i.e. as-built records. Otherwise, the delay analysis will demonstrate essential failures and may be rejected or disregarded on the basis of being hypothetical.

For instance, in Mirant Asia-Pacific Construction (Hong Kong) Limited – and – Ove Arup and Partners International Limited7 Judge Toulmin stated that:

I should refer specifically to one fundamental criticism of Mr Robinson’s evidence. Mr Lechner criticises Mr Robinson’s “final window” in his delay analysis in which he considers a hypothetical scenario assuming that the Unit 1 generator had not failed in January 1999. He says that Mr Robinson’s analysis is “a highly theoretical exercise built on unsupported assumptions which does not produce a reliable result”.

Similarly, in Civil Mining & Construction v Wiggins Island Coal Export Terminal8 , Flanagan J pointed out the importance of the facts as follows:

The flaws in Mr Abbott’s analysis may be summarised as follows:

(a) A significant number of factual assumptions made by Mr Abbott about the Contract Works are inconsistent with what actually occurred on-Site…
(b) Mr Abbott’s analysis was a theoretical exercise…

For reasons that are developed in more detail below in dealing with the Delay Claim, I generally prefer the evidence of Mr King to that of Mr Abbott.

Finally, the Society of Construction Law Delay and Disruption Protocol (SCL Protocol) states that:

Irrespective of which method of delay analysis is deployed, there is an overriding objective of ensuring that the conclusions derived from that analysis are sound from a common sense perspective9 

These two cases (and many others not provided in this article) and the SCL Protocol underline the criticality of factual substantiation and compliance with reality and common sense.

In light of the foregoing, it can be argued that facts and common sense rule the delay analysis. Therefore, any fragnet to be used in TIA should rest on the facts and comply with reality.

AACE Recommended practice provides the following source validation protocols (SVP) ‘to provide guidance in the process of assuring the validity of the source input data that forms the foundation of the various forensic schedule analysis’:

  • Baseline Schedule Selection, Validation, and Rectification (SVP 2.1);
  • As-Built Schedule Sources, Reconstruction, and Validation (SVP 2.2);
  • Schedule Updates: Validation, Rectification, and Reconstruction (SVP 2.3); and
  • Identification and Quantification of Discrete Impact Events and Issues (SVP 2.4).

These protocols (together with the delay analysis methodologies detailed therein) do not only provide guidance to the analysts but also trigger the scepticism required to interrogate the delay analysis results against the facts of the project.


In conclusion, the delay analyst should demonstrate artistic and scientific skills to turn delay events into reasonable fragnets for the analysis of delay event impacts and the entitlements of the parties.

On the one hand, pure reliance on the as-built dates that extend beyond the window cut-off dates may result in unreasonable fragnets and analysis results that are hypothetical and exaggerated.

On the other hand, fragnets deviating excessively from actual project history will undermine the credibility and reliability of the analysis.

Therefore, the delay analyst should investigate the facts and circumstances surrounding the delay events, and reflect the anticipations of the parties into the analysis through fragnets that evolve during the course of the analysis. Such a structured, robust and reasonable approach will facilitate the resolution of contentious disputes either during the course of the project or upon completion and will help parties to preserve the commercial relationship while obtaining mutually satisfactory outcomes.

A clear understanding and application of reasonability is a crucial factor for a fair and reliable delay analysis.

Hakan Arslan, Managing Consultant

About the author

Hakan Arslan is a Managing Consultant at HKA and a qualified Civil and Industrial Engineer. He has a wealth of commercial, contractual and management experience gained from working on complex buildings, infrastructure, oil and gas, and industrial projects throughout the Middle East, North Africa, and Asia regions.

Hakan has been working with HKA since 2014, preparing and/or assessing claims on HKA’s most complex and technically demanding claims assignments.


1 Reasonability is a legal term. The scale of reasonability represents a quintessential element of modern judicial systems and is particularly important in the context of international disputes and conflicts of laws issues.

2 Mirant Asia-Pacific Construction (Hong Kong) Limited – and – Ove Arup and Partners International Limited [2007] EWHC 918 (TCC).

3 Walter Lilly & Company Limited v Giles Patrick Cyril Mackay, DMW Developments Limited [2012] EWHC 1773 (TCC).
Fragnet: Fraction of a network.

4 Fragnet: Fraction of a network.

5 Mechanical Contractors Association of America (MCAA), Bulletin No.CO 3 (2011).

6 AACE International Recommended Practice No. 29R-03 Forensic Schedule Analysis.

7 Mirant Asia-Pacific Construction (Hong Kong) Limited – and – Ove Arup and Partners International Limited [2007] EWHC 918 (TCC).

8 Civil Mining & Construction Pty Ltd v Wiggins Island Coal Export Terminal Pty Ltd [2017] QSC 85.

9 The Society of Construction Law Delay and Disruption Protocol 2nd Edition (February 2017).


Follow HKA on WeChat


HKA WeChat