COBRA AUBEA 2015 Sydney, Australia 8 – 10 July 2015
RICS COBRA AUBEA 2015 The Construction, Building and Real Estate Research Conference of the Royal Institution of Chartered Surveyors The Australasian Universities’ Building Educators Association Conference
Held in Sydney, Australia in association with AUBEA, the University of Technology Sydney and University of Western Sydney
8 -10 July 2015
© RICS 2015
ISBN: 978-1-78321-071-8
Royal Institution of Chartered Surveyors Parliament Square London SW1P 3AD United Kingdom
www.rics.org/cobra
The papers in this proceeding are intended for knowledge sharing, stimulate debate, and research findings only. This publication does not necessarily represent the views of RICS, AUBEA, UTS or UWS.
APPORTIONING DELAY LIABILITY: WHICH CRITICAL PATH? Nihal Ananda Perera1 and Monty Sutrisna2 1
School of Built Environment, University of Salford, Salford, M5 4WT, UK. Department of Construction Management, School of Built Environment, Curtin University, Australia.
2
ABSTRACT Since its advent, Critical Path Method (CPM) has assumed a primary role in construction project management, particularly in planning and controlling to ensure timely completion. In addition, it has increasingly been used in resolving construction delay claims, including determining parties’ liability and resulting compensation. In its primary role, CPM enables analysts to predict the project completion date taking into account the dynamic project circumstances. However, its other role in delay claims is a more perplexed one. This is because there have been two fundamentally different approaches for defining applicable ‘critical path’. These approaches are led by two distinct theories: ‘Total Float Value’ (or Zero Float school) and ‘Longest Path’ (or Lowest Float Value school). These theories could potentially generate completely opposite outcomes for the same delay events. It is a sure recipe for disputes, if either theory is followed arbitrarily particularly in the case of concurrent delays. Thus, deciding the appropriate theory to be implemented in a particular circumstance is required to minimise the potential for disputes. This discussion is presented in this paper with a hope to promoting a broader dialogue on some considerations indispensable to choose the ‘critical path’ proper for apportioning liabilities. The ensuing discussion is largely relied on published academic works on the subject and related case law. Key Words: delays, apportioning liability, critical path.
INTRODUCTION In both US and UK courts it is now a requirement that a delay must be shown to be critical in order for awarding time or time-related damages. Here, CPM programming is key to demonstrate those events and their impacts causing critical or non-critical delays (Keane and Caletka, 2008). A delay analyst who uses CPM to undertake this task would have to select between two primary criteria for defining the ‘criticality’ of the delays. These criteria are advocated by two fundamental theories, famously known as the Longest Path (or Lowest Float Value school) and Total Float Value (or Zero Float school). Accordingly, for determining the ‘criticality’ of a delay, a delay analyst would have to decide whether all activities in the CPM based construction programme having total float less than or equal to zero are critical, or only those having the maximum negative float are critical; while the former criterion belongs to the Total Float Value theory, the latter represents the Longest Path theory. Depending on which theory is relied on, the two criteria would possibly generate completely divergent outcomes in the apportioning of liabilities and entitlement. With the intention to promoting a broader dialogue and awareness on some considerations indispensable for determining the ‘critical path’ proper, the distinct characteristics and mechanisms of these two theories are discussed here.
1 2
[email protected] [email protected]
Theories of ‘Criticality’ As noted above, the definition of ‘criticality’ can be set to either all activities on the Longest Path or alternatively, all activities with a negative float. These two definitions underpin the two fundamental theories. ‘Recommended Practice No. 29R-03 Forensic Scheduling Analysis’ (AACE International, 2011) has explicitly identified the fundamental differences between these theories. Understanding of these differences is of vital importance for determining the ‘critical path’ proper. A ‘negative’ float is formed in CPM through imposing a ‘constraint’. Once a schedule is constrained to its completion date that alters one of the basic rules of CPM scheduling (O’Brian and Plotnick, 2005). In doing so the late finish of the last activity (presuming it is tied to a contract completion date) becomes equal to the early finish of that activity; accordingly, if that last activity is delayed beyond that late finish date, the calculation of the ‘Total Float’ will be a negative value. Thus, the network model may display negative float values indicating how many days the schedule activity is behind schedule or the contract completion date. The general norm as to a CPM programme’s critical path is that it is the Longest Path of logically connected activities which, when the individual time durations of each activity are added, equals the overall duration of the project (AACE International, 2011). However, at any given time there may also be multiple paths parallel to this longest critical path (Wickwire et al., 2003). Albeit this Longest Path is perfectly capable of predicting the project completion date at a given time, it will not readily reveal the existence of delays in the other subordinate paths running parallel to the Longest Path. The delays in these subordinate paths are critical to the prevailing contract completion date but would have no impact on the predicted project completion date which is set by the Longest Path. Where two causes of delay are of different causative potency, according to the Longest Path theory the longer delay is then regarded as the effective cause and the shorter or the subordinate one as the ineffective. In other words, the minor cause is treated as if it were not causative at all (Marrin, 2002). Under the Longest Path theory, if an activity has a float with respect to the Longest Path in excess of a given delay, it can absorb that delay, and therefore no time extension will be required relative to the existing project completion date. In this instance, a ‘residual float’ will be created by the largest negative float of the Longest (critical) Path which can absorb the shorter delays in the subordinate (critical) paths; thus, under the Longest Path theory the mere fact that an activity has a negative float with regard to a contractual milestone will not determine its criticality in so far as the project completion date is not affected. Consequently, in spite of having their own negative floats any delays in the subordinate paths are consigned to oblivion by the maximum negative float in the Longest Path and hence assumed ‘non-critical’. On the other hand, if the ‘criticality’ is determined by Total Float Value theory all activities that have negative float (i.e. one or more unit below zero total float) relative to the existing contract completion date are considered ‘critical’ (Jentzen et al., 1994; Peters, 2003). Bramble and Callahan (2000, p.11-88) explained the Total Float Value theory as “In other words, in addition to the one critical path, any path with negative float may qualify as “critical” under this approach and be available to offset a
claimed delay…..In the absence of delay to the one critical path, all other paths with negative float also would have caused delayed project completion”. It is obvious from the foregoing discussion that ‘criticality’ in context cannot be considered an absolute phenomenon but has only a relative meaning and value. Thus, the ‘criticality’ of an effect of delaying event is determined whether it is measured against the (predicted) project completion date or from the point of (prevailing) contract completion date.
Concurrent Delays and Determining ‘Criticality’ For determining the ‘criticality’ of delays, the importance of concurrent delays is relevant to one issue only. That is compensability. When all delays are critically affecting the time for completion, a party who needs to defend a claim would find the significance of a concurrent delay as it may permit him to offset the compensability claimed by the other party (Bramble and Callahan, 2000). In other words, its use is for the contracting parties trying to cancel out the compensability of one another. However, thanks to the numerous definitions floating around this already complex issue has been further complicated (Zack and Federico, 2011). Basically, the controversy exists in whether a situation where two or more delaying events (one an employer risk event and the other a contractor risk event) occur at the same time and have effects felt at the same time or a situation where two or more such events arise at different times and have effects felt at the same time should be considered for concurrent delays (SCL Protocol, 2002). This controversy on concurrent delays is inextricably linked with the approach for defining the ‘criticality’ of the delays (Peters, 2003). Here, the proponents of Longest Path approach incline to recognize ‘true concurrency’ or concurrency at the same time as the ideal situation which gives rise to concurrent delays. For example, Wickwire et al. (2003) who championed Longest Path approach for CPM delay analysis recognized the concurrent delays as a situation in which two or more delaying events from each party occurred at the same time during all or portion of the delay periods being considered; for others, to be considered as concurrent delays, the delays need not actually taking place at the same time (Ness, 2000) and concurrent delays may occur either (i) when there are two or more independent delays during the same period, or (ii) during any part of the project performance period, sequentially and not necessarily at the same time (Bramble and Callahan, 2000; Tobin, 2007; ‘SCL Protocol’, 2002). Particularly, the ‘Figure 9’ of the SCL Protocol illustrates this with sequentially occurring delaying events on two independent paths (one being the ‘longest’ and the other ‘subordinate’) and acknowledges their resulting concurrent effects of unequal potency in order for apportioning liabilities and offsetting. Notably, with the principle illustrated in ‘Figure 9’ (see ‘Figure 9’ re-produced below), SCL Protocol has unambiguously considered the ‘effect’ of the subordinate path is also critical and concurrent with the ‘effect’ of the Longest Path. This is in contrast to the orthodox Longest Path approach which considers the ‘effect’ of a subordinate path as ‘noncritical’ (hence ‘non-concurrent’), and only the path with activities having the maximum negative float (i.e. the lowest float) is critical. There appears no universally accepted legal position defining what should be considered as the meaning of ‘concurrency’ or ‘concurrent delay’ (Moran, 2014). For example, in the UK jurisdiction there are case authorities for both definitions. Thus, in
‘Malmaison’3 it seems the judgment considered the concurrent causes occurring at the same time only. The ‘Malmaison’ test was followed by several other cases too. On the other hand, the second definition above seems corroborating with the judgment in The Royal Brompton Hospital4; there, the Judge Seymour considered an argument that two delays happening at different times were concurrent delays. Occasionally a similar position can be seen in the US jurisdiction as well. For example, in Raymond Construction of Africa, Ltd.5 the court determined that three consecutive delays were concurrent; similarly, in Williams Enterprises, Inc6. the federal court determined consecutive delays as concurrent delays. The foregoing calls for considering the issue of defining ‘concurrency’ not so dogmatically. Perhaps, unless expressed otherwise in a contract, it may be safer to consider that both ‘true concurrency’ (i.e. two or more delay events occur at the same time) and ‘concurrent effects’ (i.e. two or more delay events arise at different times but the effects of them are felt at the same time) should have equal effectiveness as considered by SCL Protocol core principles (ref. item 1.4, and ‘Figure 9’); it may not only corroborate with the mainstream legal position to an extent, but also offer a more equitable approach to agree on for defining the ‘criticality’ of delays. Which Critical Path? AACE International (2011) and Livengood and Peters (2008) pointed out that as all forensic delay methodologies provide for extensions of contract time on the critical path, only the definition for critical path is of utmost importance. Thus, the allimportant question that a delay analyst may face in determining the ‘criticality’ of the delays is whether all activities having total float less than or equal to zero are critical (Total Float theory), or only those having the maximum negative float are critical (Longest Path theory). The answer to this question would also decide which position should be adopted in defining ‘concurrent delays’ for setting-off compensability. The SCL Protocol’s ‘Figure 9’ can be utilised here to illustrate the fundamental difference of outcomes between these two theories. Two scenarios are considered for this, using Path 1 and Path 2 as shown under ‘After Change Event’ of ‘Figure 9’: For the first scenario, it is assumed that the delay in Path 1 (the Longest Path with 7 days of negative float) has been caused by an employer’s excusable and compensable delay and the delay in Path 2 (the subordinate path with 4 days of negative float) is caused by a contractor’s non-excusable delay. According to Longest Path theory, the contractor may consume a negative float (created by the longest delay of Path 1) up to the (predicted) project completion date as long as the contractor’s non-excusable delay in subordinate path (Path 2) finished earlier than the employer’s longest delay. Consequently, the contractor’s delay is considered non-critical relative to the Project Completion Date which is set by the Longest Path; in turn, he is entitled to recovery of extended overhead expenses with extension of time for the entire length (7 days) of the Longest Path delay. However, if Total Float theory is applied here, both Path 1 and Path 2 delays are considered critical relative to the prevailing Contract Completion
3
Henry Boot Construction (UK) Ltd V Malmaison Hotel (Manchester) Ltd [1999] 70 Con L.R. 32 The Royal Brompton Hospital NHS Trust v Hammond And Others [2001] 76 Con L.R. 148 5 Raymond Construction of Africa, Ltd. v United States, [1969] 411 F2.D 1227 Ct. CI 6 Williams Enterprises, Inc. v Strait Manufacturing & Welding, Inc., [1990] 728f. Supp.12 D.D.C. 4
Date and concurrent as well to the extent of their overlapping. Consequently, although the contractor is entitled to extension of time for the entire length of the Longest Path delay his entitlement to recovery of extended overhead expenses is limited for the nonconcurrent (3 days) period only; in this instance the employer does not have to compensate for extended overhead costs for the entire (7 days) Longest Path delay period of employer’s compensable delay. For the second scenario the delay causes are reversed: now the assumption is the delay in Path 1 (the Longest Path) has been caused by a contractor’s non-excusable delay while the delay in Path 2 (the subordinate path) is caused by an employer’s excusable and compensable delay. According to Longest Path theory, the employer may consume a negative float (created by the longest delay of Path 1) up to the (predicted) project completion date as long as the employer’s delay in subordinate path (Path 2) finished earlier than the contractor’s longest delay. Consequently, the employer’s delay is non-critical relative to the Project Completion Date which is set by the Longest Path, and in turn he is entitled to claim Liquidated Damages for the entire length (7 days) of the Longest Path delay. However, if Total Float theory is considered, both delays are treated critical relative to the prevailing Contract Completion Date and concurrent as well to the extent of their overlapping. As a result, the contractor is entitled to extension of time for the overlapping period (4 days), and the employer is entitled to Liquidated Damages for the non-concurrent (3 days) period only; in this outcome the employer is not entitled to Liquidated Damages for the entire (7 days) Longest Path delay period of contractor’s culpable delay.
The SCL Protocol ‘Figure 9’ (Source – SCL Protocol, 2002) The foregoing illustrates a marked contrast of outcomes of entitlement and the liability of each party based on which approach/theory is used to define the ‘criticality’ for offsetting the compensability. In many jurisdictions, including the US and the UK, the CPM based forensic schedule analysis is almost mandatory in delay claims resolution. Nevertheless, in the absence of clear contract terms, the legal position as to which approach or theory to be followed to define ‘criticality’ for offsetting the compensability is not clear. Even in the US where CPM based forensic schedule analysis is given more prominence than in any other jurisdiction a debate is continuing on the issue. A landmark case in this
regard is Santa Fe, Inc.7. In this case, there was a Liquidated Damages (LD) clause in the contract and as the contractor (Santa Fe) delayed, LDs were imposed. The contractor appealed for a remission of LDs and sought extension of time. An express provision of the ‘Santa Fe’ contract entitled ‘Adjustment of Contract Completion Time’ stated as follows: “Actual delays in activities which, according to the computer-produced calendar-dated schedule, do not affect the extended and predicted contract completion dates shown by the critical path in the network will not be the basis for a change to the contract completion date”. The above contractual provision was so clear that the ‘criticality’ had to be determined by the impact on the ‘extended and predicted’ project completion date only and not otherwise. Thus, it was decided that the contract required to use Longest Path approach to measure ‘criticality’; in view of this express provision the tribunal rejected the contractor’s argument “…any work sequence or CPM path which runs past its contractually required completion date to be critical and any delays on those work sequences to be on the critical path” and held “there is still a critical path represented by the negative slack activities with the highest numerical designation (for example -180 days versus -50 days), The activity chain representing the highest negative slack (for example, the -180 days) represents the longest chain of activities through the project in terms of time” (Wickwire et al., 2003, p.376). It could have been interesting to see how the ‘Santa Fe’ would have been decided if that express provision was not present in the contract, but the court did not make any comment on that aspect. On the other hand, in Fischbach & Moore8 case, though the CPM was used, the tribunal considered the other critical paths as well together with the longest path for determining the criticality of delays. It is interesting to compare the stated positions between the two famous Protocols published on either side of the Atlantic (i.e. US based ‘RP No. 29R-03’ of AACE International and the UK based ‘SCL Protocol’) on choosing the ‘critical path’ for determining ‘criticality’. On one hand, the RP No. 29R-03 uses the Longest Path theory “as the valid criterion for criticality where negative float is shown” (AACE International, 2011, p.114). However, on the same page it acknowledges the relative correctness of the two theories stating “which one is correct depends on which principles are considered”. On the other hand, the SCL Protocol (2002) takes the position “Unless there is express provision to the contrary in the contract...an EOT should only be granted to the extent that the Employer Delay is predicted to reduce to below zero the total float on the activity path affected by the Employer Delay” (p.13) and “Employer Delay to Completion and Contractor Delay to Completion, both of which mean delay to a contract completion date” (p.55).[Emphasis added]; the SCL Protocol’s ‘Figure 9’ (reproduced above) clearly illustrates this position acknowledging that all activities (regardless whether in the longest or the subordinate paths) having total float less than or equal to zero are critical and the ‘criticality’ is measured against the contract 7
Santa Fe, Inc., [1984] VABCA Nos. 1943-1946.
8
ASBCA No. 18.146,77-1 BCA '/112,300 (1976).
completion date and not the project completion date; this clearly corroborates with the position of Total Float theory. CONCLUSIONS In the foregoing discussions, the paper has identified the characteristics and mechanisms of Longest Path theory and Total Float theory as they are used for determining the ‘critical path’ proper for apportioning liabilities. It has also explored some essential considerations as regards ‘when to follow which theory’, particularly in the backdrop of ‘concurrency’ of delays. The two theories clearly show that their characteristics and mechanisms are different and mutually exclusive. Having appraised the appropriate circumstances for the use of each theory the main conclusions of the paper are as follows: Between the two theories, there is no singularly correct definition for ‘criticality’; the ‘criticality’ can be set to either all activities on the Longest Path only (represented by Longest Path theory or Lowest Float Value school) or all activities with a negative float (represented by Total Float Value theory or Zero Float school); which one is correct depends on which principles are considered. Although the delaying events being analyzed are the same, there could be completely opposite outcomes of entitlement and liability of each party depending on which theory/approach is used to define the ‘criticality’ for offsetting the compensability. If the contractual considerations require only the CPM principles to be used and consider only the path showing the maximum negative float should be ‘critical’ (to the Project Completion Date) for applying EOT or LD, the Longest Path theory is correct; if the contractual considerations require to consider all paths showing negative floats (albeit not equally) and EOT or LD be applied to the extent of such negative floats are ‘critical’ to the (prevailing) Contract Completion Date, the Total Float Value theory is correct. Thus, which theory to be followed is best agreed in the terms of contract, prior to the onset of delays. Their distinctive characteristics and potential contrasting results preclude mixed use of the two approaches/theories; for example proponents of SCL Protocol principles (which are based on ‘concurrent effects of sequential delays’ and treating both ‘longest’ and ‘subordinate’ delays as critical relative to the prevailing Contract Completion Date) may find using Longest Path approach to define the ‘criticality’ for offsetting compensability is paradoxical and a mismatch to SCL Principles. Liability in contracts will depend upon the terms of the contract and the intention of the parties (Atkinson, 2007). Thus, in conclusion, in deciding the appropriate theory to define the ‘criticality’ for offsetting compensability, the first port of call should be the express contractual considerations. In absence of such considerations, the best alternative is to make effort to decide it through mutual consent between the parties, without being arbitrary or biased towards either theory. It is hoped that the above submission may promote a broader dialogue in the delay analysis domain wherein more objective and less bias approach is conscientiously required for determining the ‘critical path’ proper as to offsetting compensability.
REFERENCES AACE International (2011) ‘Recommended Practice No. 29R-03 Forensic Scheduling Analysis, Available: http://www.aacei.org. (Accessed: 28 July 2011). Atkinson, D. (2007) Causation in Construction Law – Principles and Methods of Analysis, London: Daniel Atkinson Ltd. Bramble, B. B. and Callahan, M.T. (2000) Construction Delay Claims. New York: Aspen Publishers. Jentzen, G. H., Spittler, P. and Ponce de Leon, G. (1994) ‘Responsibility for Delays after the Expiration of the Contract Time’, 1994 AACE International Transactions [CD Rom]. WV: AACE International, Morgantown. Keane, P.J. and Caletka, A.F. (2008) Delay Analysis in Construction Contracts. Oxford: Wiley-Blackwell. Livengood, J. C. and Peters, T.F. (2008) ‘The Great Debate: Concurrency vs. Pacing, Slaying The Two-Headed Dragon’, AACE International Transactions [CD Rom], pp cdr.06.1 – cdr.06.17. Marrin, J. QC. (2002) ‘Concurrent Delay’, Available: http://www.scl.org.uk. (Accessed: 28 July 2008) Moran, V. QC. (2014) ‘Causation in Construction Law: The Demise of the ‘Dominant Cause’ Test?’ Available: http://www.scl.org.uk. (Accessed: 8 December 2014) Ness, A.D. (2000) When the Going Gets Tough—Analyzing Concurrent Delays, Available: http://www.constructionweblinks.com. (Accessed: 30 June 2010) O’Brian, J.J. and Plotnick, F.L. (2005) CPM in Construction Management. New York: McGrow-Hill. Peters, T.F. (2003) ‘Dissecting the Doctrine of Concurrent Delay’, AACE International Transactions [CD Rom], pp cdr.01.1 – cdr.01.8. SCL Protocol (2002) The Society of Construction Law - Delay and Disruption Protocol, Available: http://www.scl.org.uk or http://www.eotProtocol.com. (Accessed: 10 September 2007) Stumpf, G.R. (2000) ‘Schedule Delay Analysis’, Cost Engineering, 42(7), pp. 32-43. Tobin, P. (2007) ‘Concurrent and Sequential Cause of Delay’, The International Construction Law Review, 24(part 2), pp. 143-167. Wickwire, J.M., Driscoll, T.J., Hurlbut, S.B. and Hillman, S.B. (2003) Construction Scheduling: Preparation, Liability and Claims, 2nd ed. London: Aspen Publishers Zack, J.G. Jr. and Federico, E.R. (2011) ‘Concurrent Delay – The Owner’s Newest Defense’, Navigant Construction Forum, Available: http://www.navigant.com/~/ media/www/site/insights/construction/concurrent-delay-construction.ashx. (Accessed: 26 June 2012)
