DSpace at VNU: Analysis of Adverse Weather for Excusable Delays

Analysis of Adverse Weather for Excusable Delays
Long D. Nguyen1; Jax Kneppers2; Borja García de Soto, P.E., A.M.ASCE3; and William Ibbs, M.ASCE4
Abstract: Severe weather conditions can be disruptive to construction. Contractors typically obtain time extensions for weather days
beyond normal conditions. However, contracting parties often dispute the extent of weather-related time extensions. Typical industry
contracts may overlook many important points that can provide an acceptable resolution. This paper classifies seven factors causing
discrepancies in analysis of adverse weather for time extensions; namely, the definition of normal weather, weather thresholds, type of
work, lingering days, criteria for lost days, lost days equivalent due to lost productivity, and work days lost versus calendar days lost. An
analysis of an actual weather-caused delay claim illustrates the impacts of those factors on the outcomes of the analysis. A contract should
define anticipated weather delay days and their lingering days and provide threshold values for weather parameters to differentiate
between predictable and unpredictable severe weather. The contract should clearly define how a time extension is granted in calendar days
as a result of work days lost, and also address how a time extension is granted due to inefficiency caused by unusually severe weather.
Future research may provide an appropriate mechanism for analyzing equivalent lost days to account for lost productivity.
DOI: 10.1061/͑ASCE͒CO.1943-7862.0000242
CE Database subject headings: Claims; Weather; Delay time; Contracts; Construction management.
Author keywords: Claims; Weather; Delay time; Contracts; Construction management.

Introduction
Construction productivity is typically exposed to and contingent
on weather conditions. These conditions are local, seasonal, and
sometimes unusual. Inclement weather conditions may result in
project delays, disruptions, and possibly disputes between the
project parties. Many trades ͑e.g., earthwork, concrete, structural
frame, roof, landscape͒ are readily affected by unexpectedly severe weather while other trades ͑e.g., ceiling, carpet͒ may be not.
Hurricane Katrina and its aftermath not only delayed construction
projects regionally but nationally and even globally due to the
shortage of construction materials and construction equipment.
Thus, owners, general contractors, and trade contractors alike face
risks associated with weather conditions. The project parties respond to such risks by adopting strategies which include: contingency, purchasing insurance, sharing or shifting risk to other
parties through the contract terms and conditions, to name a few.
1
Lecturer, Faculty of Civil Engineering, Ho Chi Minh City Univ. of

Technology, Ho Chi Minh City, Vietnam; formerly, Construction Consultant, Jax Kneppers Associates, Inc. ͑JKA͒, 2125 Ygnacio Valley Rd., Suite
200, Walnut Creek, CA 94598 ͑corresponding author͒. E-mail:

2
President, Jax Kneppers Associates, Inc. ͑JKA͒, 2125 Ygnacio Valley
Rd., Suite 200, Walnut Creek, CA 94598. E-mail: a-jax@jaxkneppers.
com
3
Construction Consultant, Jax Kneppers Associates, Inc. ͑JKA͒, 2125
Ygnacio Valley Rd., Suite 200, Walnut Creek, CA 94598. E-mail:

4
Professor of Construction Management, Dept. of Civil and Environmental Engineering, Univ. of California, Berkeley, CA 94720; and President, The Ibbs Consulting Group, Inc. E-mail:
Note. This manuscript was submitted on January 8, 2009; approved
on May 26, 2010; published online on June 5, 2010. Discussion period
open until May 1, 2011; separate discussions must be submitted for individual papers. This paper is part of the Journal of Construction Engineering and Management, Vol. 136, No. 12, December 1, 2010.
©ASCE, ISSN 0733-9364/2010/12-1258–1267/$25.00.

Project owners generally allocate the risks of weather-related
delays through contract provisions such as “weather,” “default,”
and “force majeure” clauses. The common rule is to grant an
extension of time to a contractor for a delay caused by abnormally
adverse weather conditions ͑Kartam 1999͒. Such delay is normally excusable and noncompensable. How “abnormal” is defined affects the amount of time extension, especially if an
extremely severe weather day, for instance, is actually an excusable delay. Weather was the second leading cause of the 24 claims
investigated in Canada ͑Semple et al. 1994͒. It was the most
frequent reason for delay claims in 57 surveyed projects, presumably built in Canada ͑Yogeswaran et al. 1998͒.
This paper presents factors affecting the analysis of inclement
weather delays in construction projects. A disputed project is used
as a case study. Though the dispute involved several issues such
as project delays, achievement of substantial completion, and

construction defects, this paper concentrates on the different
views and analyses of adverse weather delays for the same contract. This paper also attempts to determine how these factors
should be addressed in construction contracts to reduce different
technical interpretations of abnormal weather conditions and, subsequently, different analyses and results.
A major contribution of this paper is that it expands the analysis of adverse weather for excusable delays. Thus, this research
helps improve the reliability of the analysis and hence makes
extensions of time more agreeable between project parties. Legal
interpretations of weather delay related provisions are beyond the
scope of this paper.

Related Work
Extensive research has been conducted on many facets of the
impacts of weather conditions on construction projects. Weather
is a crucial factor causing project delays and cost overruns ͑Baldwin et al. 1971; Laufer and Cohenca 1990; El-Razek et al. 2008͒.

1258 / JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / DECEMBER 2010

Downloaded 18 May 2012 to 130.220.71.20. Redistribution subject to ASCE license or copyright. Visit


Table 1. Average Number of Adverse Rain Delays for Santa Rosa, California
January
12

February

March

April

May

June

July

August

September

October

November

December

10

10

6

4

1

1

1

1

4

8

10

Other research has quantified the impacts of severe weather on
productivity ͑Grimm and Wagner 1974; Thomas et al. 1999; ElRayes and Moselhi 2001͒. Many studies have proposed better
practices in dealing with weather risks to construction activities
and trades, project schedules, or project costs, etc. ͑Maunder et al.
1971; Benjamin and Greenwald 1973; Smith and Hancher 1989;
Moselhi et al. 1997; Connors 2003; Xi et al. 2005; Chan and Au
2008͒. Several studies have investigated how weather is specified
in construction contracts ͑Hinze and Couey 1989͒ and is ruled on
litigation cases ͑Finke 1990͒. However, few studies address the
issue of differentiating unusually ͑unpredictably͒ inclement
weather from usually ͑predictably͒ inclement weather, because
this is a principle for classifying weather delays in typical construction contracts. The differentiation is actually not easy though.
It has challenged the project practitioners and claims analysts, and
is subject to speculation and manipulation.

Prerequisites for Time Extensions due to Weather
When encountering unusually severe weather, a contractor is normally entitled to obtain additional time but not additional money.
Specifically, a severe weather condition is an excusable delay if it:
͑1͒ meets contract requirements; ͑2͒ is not foreseeable; and ͑3͒
impacts the critical path. The first and the last principles seem
straightforward. The contractor has to notify the owner and/or
other authorized parties per contract requirements if a delay occurs. For instance, in Handex ͑2005͒, the Court states “we believe

there was sufficient evidence of an ‘abnormal weather condition’
as described in Handex’s weather logs and data to give the issue
to the jury. However, ͓…͔ the contract clearly states that claims
for more time due to ‘abnormal weather conditions’ would only
be considered when brought ‘within the Contract Times’ ͓…͔.
Their request, made 10 July 2001, was well outside of the contract time.” A thoughtful delay analysis would answer if a weather
delay is critical for the third principle, which is another focus of
previous research ͑Smith and Hancher 1989; Moselhi et al. 1997;
Moselhi and El-Rayes 2002͒. Schedule delay analysis has also
been improved recently ͑Hegazy and Menesi 2008; Nguyen and
Ibbs 2008͒. The second principle is more ambiguous and disputable as discussed below.
In the construction industry, “foreseeability is the standard applied to many excusable delays by courts interpreting and applying any clauses … Foreseeable delays that are often deemed nonexcusable for the contractor include delays due to normal weather
conditions” ͑Bramble and Callahan 2000͒. Thus, a contractor cannot allege that all severe weather conditions are excusable delays.
Severe weather conditions in excess of the total expected or foreseeable for the duration of the project for a given location may be
the basis for a contract time extension if the contractor can demonstrate that the unexpected adverse weather delayed activities on
the critical path.
In the American Institute of Architects ͑AIA͒ Document A2012007 ͑American Institute of Architects 2007͒, “General Conditions of the Contract for Construction” ͑hereafter referred to as
AIA A201͒, which is widely used in the U.S. construction industry, subsection 15.1.5.2 says “If adverse weather conditions are

the basis for a Claim for additional time, such Claim shall be
documented by data substantiating that weather conditions were
abnormal for the period of time, could not have been reasonably
anticipated and had an adverse effect on the scheduled construction.” However, abnormal or “not … reasonably anticipated” is
not defined. This may create speculation and dispute among the
project parties and schedule consultants involved in the claims
and disputes. Results derived by one party/schedule consultant
cannot be reassembled by the other. Consequently, an equitable
settlement can be difficult to achieve.

Factors Affecting Weather Delay Analysis

Various factors influence analysis of severe weather and its results. This research systematizes these factors as discussed below.
Definition of Normal Weather
Normally severe weather conditions should be incorporated in all
construction contracts and the contract should differentiate between normally and abnormally severe weather conditions. A
clear definition of “normally anticipated weather” should exist to
avoid any ambiguity ͑Hinze and Couey 1989͒. Some owners and
contractors now provide the number of anticipated weather days
in their contracts. In other words, contracts specify the average
͑normal͒ number of adverse rain days for the project location.
Table 1 presents such an example of the number of adverse rain
days included in the project manual of a public contract in Santa
Rosa, California. The basis for deriving the monthly anticipated
adverse weather delay is typically not provided though. That is,
contracts may specify the numbers of normal weather delay days
but do not describe where they come from. Disputes may arise as
to what constitutes an unusually severe weather when the basis
for comparison is unclear. Though presenting that the precipitation was greater than the 20-year average, a contractor’s claim
failed because the precipitation, as shown by the owner’s expert,
was lower than the past 5-year or 24-year average ͑McDevitt
1989, cited in Bramble and Callahan 2000͒.
With regulation number ER 415-1-15, the U.S. Army Corps of
Engineers ͓U.S. Army Corps of Engineers ͑USACE͒ 1989͔ provided the methodology for construction time extensions for
weather-related delays for its contracts. However, guidance for
this methodology has changed over time and acknowledged in its
construction bulletins ͓U.S. Army Corps of Engineers ͑USACE͒
1996, 2008͔. The time extension granted in calendar days, for
instance, is determined differently. U.S. Army Corps of Engineers
͑USACE͒ ͑1996͒ determined the time extension as a product of
the work day delay and “a ratio of seven calendar days per week
to the number of scheduled work days per week” whereas U.S.

Army Corps of Engineers ͑USACE͒ ͑2008͒ instructed that it
“should be based on the new calculated early finish date as a
result of the added unusually severe weather frag-net.” Inconsistent application of ER 41-1-15 is also recognized by the unreasonable evaluation of what constitutes usually severe and
unusually severe weather ͓U.S. Army Corps of Engineers
͑USACE͒ 2008͔.

JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / DECEMBER 2010 / 1259

Downloaded 18 May 2012 to 130.220.71.20. Redistribution subject to ASCE license or copyright. Visit


The definition of normally severe weather becomes more awkward when contracts do not provide the number of anticipated
weather delay days. Contract documents have an unclear description of normal weather ͑Smith and Hancher 1989͒. In many contracts, wind, temperature, precipitation weather parameters and
weather statistics such as the length of historical weather data,
mean versus median of the data, etc., are unavailable to help
establish the anticipated weather. In addition, unusually severe
weather days are different between weather analyses based on a
month-by-month basis versus those based on the whole duration
of the contract work. For instance, the Interior Board of Contract
Appeals once held that the delay caused by unusually inclement
weather in one month can offset the unusually good weather in
another month ͑Bramble and Callahan 2000͒. Unless contracts
account for and provide specific definitions for these items, time
extensions will be difficult to assess due to a multitude of ways to
define normally severe weather coupled with the many possible
answers regarding the foreseeability or unforeseeability of
weather-related delays.
Weather Thresholds
Inclusion of weather thresholds in construction contracts is important because they determine what qualifies as a weather delay
day. An example would be that “work must be completely

stopped if temperatures are below −10° F ͑−23.3° C͒ or above
110° F ͑43.3° C͒.” This is because efficient construction operations are difficult to achieve with those temperatures ͑Koehn and
Brown 1985͒. This factor is somewhat related to the definition of
normally adverse weather. Intensity of rainfall, type of construction operation, and on-site drying conditions are three key factors
affecting weather-related delays and disruptions ͑El-Rayes and
Moselhi 2001͒. Intensity of rainfall is associated with weather
thresholds while the last two factors are associated with the next
two dimensions, namely type of work and lingering days.
Different threshold values for a given weather condition result
in different number of normally and abnormally severe weather
days for the same actual conditions during the same timeframe.
The Boards of Contract Appeals ͑BCA͒ sometimes determine inconsistent threshold values. The National Aeronautics and Space
Administration ͑NASA͒ BCA once viewed the severe weather as
at least a trace ͑less than 0.01 in. or 0.254 mm͒ of rainfall while
the Armed Services BCA ͑ASBCA͒ in a case rejected this threshold value and used a daily severity of 0.5 in. of rainfall ͑Finke
1990͒. Threshold values also depend on trades and natural and
social factors. For reference only, effective on October 14, 1993
NASA discontinued its BCA and let its contract appeals be decided by ASBCA ͑Worthington and Goldsman 1998͒.
Some contracts already determine threshold values for adverse
weather. The Tennessee Department of Finance and Administration ͓Tennessee Department of Finance and Administration
͑TennDFA͒ 2007͔, for example, establishes threshold values as
0.10 in. ͑2.54 mm͒ for precipitation ͑rain, snow, or ice͒, temperatures not above that required for the work of the day, and sustained wind above 25 mi/h ͑40 km/h͒ for its capital projects. No
single threshold value for temperature exists because the characteristics of a particular work decide the range of temperature that
the work can be performed. For instance, the temperature threshold of asphalt pavement is lower than that of concrete pavement.
With prima facie thresholds, the contractor only needs to show
that the actual conditions exceeded the thresholds ͑Xi et al. 2005͒.
However, a weather day should be decided based on the combination of the threshold values, type of work affected, and com-

mon sense ͑McDonald 2000͒. Kenner et al. ͑1998͒ recommended
that if two parties do not agree whether the weather conditions

support working or nonworking, then the decision will be based
on the comparisons of the threshold values and actual weather
data of the day in question.
Type of Work
Type of work may also define the weather impact. Certain activities are particularly sensitive to weather conditions ͑Attanasi et al.
1973͒. In addition, several factors ͑e.g., duration, exposure, material characteristics, equipment used, level of protection provided͒ for each activity influence the activity’s sensitivity to wind,
temperature, and precipitation ͑Smith and Hancher 1989͒. A
project’s weather sensitivity also depends on the phase of construction. The initial site phase of building construction, for example, is usually more sensitive to weather than the dried-in
phase ͑McDonald 2000͒. Consequently, the allocation of monthly
anticipated adverse weather delay in contracts is not sufficient.
The scheduled activities and their sequence should be known before the number of monthly weather days has been determined
because the weather impact is associated with the type of activities being performed ͑Xi et al. 2005͒.
Lingering Days
The effect of weather impact on a project may not end when the
severe weather concludes. It may extend beyond actual severe
weather periods due to site preparation, before/after snowy conditions, drying time for soils, and accumulated water on the site
͑Finke 1990͒. “Dry-out” or “mud” days are also used to describe
lingering days. McDonald ͑2000͒ categorized weather days as
weather event day, mud day, and rework or impact day. In at least
one instance a contractor was entitled to a time extension due to
unplanned work performed by the contractor and necessitated due
to unusually severe weather conditions ͑Finke 1990͒.
Even when specifying the number of monthly normally adverse weather days, a contract is still ambiguous if it does not
determine how to deal with lingering days. Contracts should provide a definition for lingering days and clearly spell out whether
lingering days are included in the monthly normal weather days.
In Tennessee Department of Finance and Administration
͑TennDFA͒ ͑2007͒, adverse weather may include dry-out or mud
days “at a rate no greater than 1 make-up day for each day or
consecutive days of beyond the standard baseline that total 1.0
inch or more, liquid measure….” El-Rayes and Moselhi ͑2001͒

investigated the lingering effects of rainfall for four highway construction activities, namely earthmoving, base courses, drainage
layers, and paving.
Differentiating between lingering days caused by predictable
and unpredictable severe weather conditions is also important.
This is because two options are available. Dry-out days can be
counted as delay days prior to the analysis of usually or unusually
severe weather. Alternatively, any dry-out day caused by usually
severe weather can be considered usually severe. Consequently,
the monthly anticipated adverse weather days must be defined
differently under these two options. Tennessee Department of Finance and Administration ͑TennDFA͒ ͑2007͒ followed the second
option by specifying that dry-out or mud” days are not counted as
a weather delay day until the anticipated adverse weather delay is
exceeded. When a month-by-month analysis for adverse weather
is used with the second option, the contract should define how to
deal with lingering days that “roll over” into the next month.

1260 / JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / DECEMBER 2010

Downloaded 18 May 2012 to 130.220.71.20. Redistribution subject to ASCE license or copyright. Visit


1 2 3 4

March 2008
April 2008
5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 1 2 3 4

R R R R

R R D

R R

10 rain (R) days; 1 "dry-out" (D) day

R R D

R

R R D

3 rain (R) days, 1 "dry-out" (D) day

1 "Roll-over"

Fig. 1. Rain and lingering days in March 2008

Fig. 1 illustrates the issues with lingering effects. The shaded
dates are weekends and holidays. The data shown come from
contemporaneous project records in the aforementioned public
contract in Santa Rosa, Calif. ͑Table 1͒. The contractor was not
able to work on any rain or dry-out days. Month-by-month analysis is assumed to be acceptable. The average number of adverse
rain delays for March and April are 10 and 6 days, respectively
͑Table 1͒.
Determining the number of abnormally adverse weather days
in March 2008 is not easily answered. Whether the numbers
͑Table 1͒ include only anticipated rain days or a combination of
anticipated rain and dry-out days is unclear. If dry-out days are
included, the answer is simple. Any delay, regardless of rain or
dry-out days, beyond 10 days is unusually severe. That is, the

contractor would experience four unusually severe weather delay
days ͑dates 20, 21, 26, and 31͒ in March 2008. If dry-out days are
not included, the answer is much more difficult to ascertain. Table
1 would be unhelpful if dry-out days are counted prior to differentiating unusually severe from usually severe. Table 1 would still
be useful if only dry-out days caused by unusual rains are unusual
as in Tennessee Department of Finance and Administration
͑TennDFA͒ ͑2007͒. The contractor would experience three unusually severe weather days ͑dates 21, 26, and 31͒ in this scenario.
However, the anticipated challenge is the dry-out day that rolls
over into April 2008 ͑date 1͒. Unless clearly predefined in the
contract, speculation is needed to determine whether a roll-over
dry-out day is considered a normally or abnormally adverse
weather.
Criteria for Lost Days
Criteria for defining lost days due to weather events vary in contracts. In fact, many contracts do not define such criteria at all.
The percentage of absent workforce, the percentage of the day
worked, or a combination of these two is often used to define a
lost day ͑Hinze and Couey 1989͒. To qualify as actual severe
weather delay days, U.S. Army Corps of Engineers ͑USACE͒
͑1989͒ required that critical activities cannot be performed for
50% or more of the contractor’s scheduled work days. Hinze and
Couey ͑1989͒ recommended that the percentage of the day that
cannot be worked be used as the criterion for defining lost days
caused by unusually severe weather.
Lost Days Equivalent due to Lost Productivity
Adverse weather reduces labor productivity. This is especially
true for construction that is typically exposed to weather. Working
under adverse weather condition may make the performance less
efficient though may not stop the work ͑Bramble and Callahan
2000͒. Inefficiency or decreased productivity due to adverse
weather can contribute to schedule delay. That is, adverse weather

not only delays a project due to such direct lost days as weather
event, dry-out, mud, “rework,” and “unplanned work” days, it

also prolongs schedule activities due to lost productivity or inefficiency. That, in turn, may delay the overall project schedule. In
many circumstances, a contractor is not entitled to compensation
incurred by lost labor productivity caused by unusually severe
weather. That is, only lost time associated with unusually adverse
weather can be recovered.
This paper uses the term “lost days equivalent” to describe the
project delay due to productivity losses caused by unusually severe weather. Contractors may recover such lost days if the prerequisites for time extension are met. However, in most cases
contractors fail to acknowledge lost days equivalent in their
analyses and subsequent requests for time extensions. A possible
explanation for this oversight is that indirect weather impact is
probably not realized at the time of occurrence but several months
later ͑Hinze and Couey 1989͒. A time extension was granted in a
General Services BCA’s decision due to reduced productivity despite the fact that the contractor worked during the periods of
unusually severe weather ͑Finke 1990͒. McDonald ͑2000͒ recommended that productivity impacts caused by adverse weather
must be incorporated into the schedule.
Work Days Lost versus Calendar Days Lost
Project durations are typically expressed in calendar days. Similarly, time extensions are typically requested and granted in calendar days. On the other hand, adverse weather conditions are
usually expressed in work days. Lost days due to weather therefore have to be converted to calendar days. In typical practice the
number of lost days in calendar days equals the number of work
days lost times a conversion factor, 7/5, if five work days per
week are used as in U.S. Army Corps of Engineers ͑USACE͒
͑1996͒. This method may not be appropriate because the number
of available work days is different from month to month in a
calendar year.
In one case the U.S. Court of Claims held that the number of
winter days granted to the contractor was not equal to the number
of excusable delay days encountered in the summer season

͑Bramble and Callahan 2000͒. As a result, the seasonal weighting
of days may be used as a better mean for converting from work
days to calendar days ͑Hinze and Couey 1989͒.
As an example, a seasonal weighting approach is applied to
the above case ͑Table 1, Fig. 1͒. This example assumes that the
number of monthly anticipated severe weather ͑Table 1͒ includes
dry-out days and that all lost days due to weather during March
2008 ͑Fig. 1͒ are critical. If the probability of occurrence of an
adverse rain day is the same for any day of the month, the number
of anticipated rain days that occurred in weekdays during March
would be approximately 7 days ͑10 anticipated rain days multiplied by 21 work days and divided by 31 calendar days͒. Thus,
there are 14 available work days in March 2008 ͑31 calendar days
minus 10 weekend days minus 7 anticipated rain days in weekdays͒. As a result, the four unusually severe weather days in

JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / DECEMBER 2010 / 1261

Downloaded 18 May 2012 to 130.220.71.20. Redistribution subject to ASCE license or copyright. Visit


Case Study
This research chooses case study over other research methods,
such as surveys or analysis of archival information. Case study
methodology helps us understand insight into the impacts of the
factors presented above on analysis of adverse weather for excusable delays in a real-life context.
Project Description
The case study presented in this paper is two one-story tilt-up
office buildings located in Livermore, California ͑hereafter refer
to as Project ABC͒. The original contract price and project duration were approximately $5 million and 265 calendar days, respectively. Eleven changer orders were issued but caused little
change in the contract dollar amount.
The owner and the general contractor drafted the agreement

themselves, using commonplace contract conditions. The contract
did not determine the number of normally adverse rain days or
mud/dry-out days, nor did it include the weather threshold to be
used. AIA A201 ͑American Institute of Architects 1997͒ General
Conditions were part of the contract. Liquidated damages
amounted to $5,000 per calendar day ͑$5,000/day͒. The contract
specified that inclement weather conditions beyond normal were
excluded. Although there were some disputes between the parties
as to what was included in the contract, they agreed that the
contractor was eligible for time extensions due to unusually severe weather. The contractor complied with contract notice requirements.
Work started in mid-October of 2005 and required substantial
completion by mid-July2006. The buildings were not dried-in
until mid-April of 2006. That is, the initial site phase was through
the 2005/2006 winter which was the rainy season in Livermore,
Calif. Critical construction activities affected by adverse weather
included, but were not limited to, slab on grade, casting and erecting tilt-up panels, erecting and welding steel columns, installation
of wood roof structures, and roofing. Due to the limited slab area,
activities for casting and erecting tilt-up panels were divided into
two stages in the as-planned schedule. The geotechnical report
revealed moderately expansive soils at this project site.

20
Rain Days (Precipitation > 0.254 mm)

March 2008 ͑dates 20, 21, 26, and 31 as explained in the section
“Lingering Days”͒ in the above case would be approximately 8.9
͑4 ϫ 31/ 14͒ calendar days using the seasonal weighting of days
compared to only 5.6 ͑4 ϫ 7 / 5͒ calendar days using the U.S.
Army Corps of Engineers ͑USACE͒ ͑1996͒ approach
The above example demonstrates that the method used for

converting from work days to calendar days plays an important
role for weather risk allocation in contracts. This is because the
conversion results are inconsistent. Nevertheless, many contracts
do not specify an appropriate conversion mechanism. Disputes
may arise if contracts do not define a method for converting from
work to calendar days when determining the number of delay
days. The U.S. Army Corps of Engineers recently instructed its
agencies to extend the performance period based on the result of
the added abnormally adverse weather “frag-net” in the schedule
͓U.S. Army Corps of Engineers ͑USACE͒ 2008͔. This is perhaps
the most appropriate and accurate approach to avoid any confusion and inconsistency in the conversion, but also the most arduous.

18
16
14
12
10
8
6
4
2
0
November

December

January

Average 1971-2000

February

March

April

Actual 2005-2006

Fig. 2. Average and actual rain days ͑precipitation Ն0.254 mm͒ in
winter in Livermore, California

Adverse Weather as a Part of Project Delay and
Dispute
The project suffered delays. The owner established that substantial completion was achieved in mid-December of 2006, and assessed liquidated damages for more than 5 months of delay. The
contractor disagreed and brought the case to arbitration. Among
other things, the contractor claimed that the delays were out of his
control and caused by abnormally adverse weather conditions and
owner-caused delays. The owner counterclaimed for contractorcaused delays and construction defects.
Adverse weather during the period between November 2005
and April 2006 was among the major issues under the dispute.
Although the owner acknowledged that the contractor was due
time extensions as a result of adverse weather conditions, an
agreeable time extension for weather was never reached because
the parties and their consultants proposed very different numbers.
The following section shows why such a situation occurred in this
dispute.
Adverse Weather Analysis
Unusually inclement weather was claimed as a cause of delays.
Adverse weather affected and ultimately delayed the project in
the winter of 2005/2006. However, whether the delay was excusable requires proper analysis. Most importantly, the type of analysis affected the number of days the contractor should be granted if

inclement weather was beyond normal conditions. Inconsistent
answers to this question were a source of the dispute.
General analysis of the rains occurring in the 2005/2006 winter was needed to check if rains were abnormal. The contract
provided neither the monthly anticipated inclement weather days
nor threshold values for precipitation. Consequently, the baseline
for comparison was difficult to establish. Figs. 2 and 3 illustrate
the numbers of normal and actual rain days in Livermore for two
precipitation thresholds: 0.254 mm ͑0.01 in.͒ and 2.54 mm ͑0.1
in.͒. Data sources were obtained from the National Oceanic and
Atmospheric Administration ͑NOAA͒ ͑2008a,b͒. This analysis ignores the effect of the period chosen for establishing the monthly
anticipated rain days. The 30-year average, from 1971 to 2000,
was selected for convenience because statistics were available
from National Oceanic and Atmospheric Administration ͑NOAA͒
͑2008a͒. Naturally, weather statistics through 2005, when the contract actually started, would be ideal, but were not available. The
contract did not endorse any timeframe for defining normally adverse rain days.

1262 / JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / DECEMBER 2010

Downloaded 18 May 2012 to 130.220.71.20. Redistribution subject to ASCE license or copyright. Visit


Rain Days (Precipitation > 2.54 mm)

16
14
12
10
8
6
4

2
0
November

December

January

Average 1971-2000

February

March

April

Actual 2005-2006

Fig. 3. Average and actual rain days ͑precipitation Ն2.54 mm͒ in
winter in Livermore, California

Analysis shows that adverse weather in the winter of 2005/
2006 was actually abnormal ͑Figs. 2 and 3͒. There were actually
more actual rain days than could have been reasonably anticipated
in December 2005 and March and April 2006. Different weather
threshold values may draw different conclusions if a particular
rain day in a particular month is abnormal. For instance, adverse
weather in January 2006 was abnormal with precipitation equal to
or greater than 0.254 mm ͑Fig. 2͒ while it was normal with precipitation equal to or greater than 2.54 mm ͑Fig. 3͒. This confirms
that threshold values define how unusual a severe weather condition is. For Project ABC the winter of 2005/2006 was categorized

as abnormal under both circumstances. Cumulative precipitations
also support that abnormality ͑Fig. 4͒. The precipitations for December 2005 and March 2006 exceeded the average precipitation
by 165 and 94%, respectively. The actual total precipitation for
2005/2006 winter ͑November 2005 to April 2006͒ was 30% more
than average.
During the course of the work the contractor properly notified
the owner inclement weather and mud days had occurred. The
owner was notified of a total of 52 weather-related lost days. The
owner reviewed and responded to the contractor’s notices if they
did not agree or required clarification. Due to a lack of supporting
information provided to the owner, a total of fifteen weatherrelated lost days were rejected by the owner. The contractor accordingly revised the delay notices and removed twelve of the
fifteen lost days rejected by the owner. The contractor still maintained his view with respect to three lost days that the owner did

160

Precipitation (mm)

140
120
100
80
60
40
20
0
November

December

January

Average 1971-2000

February

March

April

Actual 2005-2006

Fig. 4. Average and actual precipitations in winter in Livermore,
California

not agree with. That is, there were forty ͑52− 12͒ lost days ͑including three disputable lost days͒ timely notified and recorded at
the end of the contract ͑Table 2͒.
Fig. 5 illustrates the flowchart for analyzing adverse weather
and Table 2 summarizes the analysis. Columns ͑2͒ and ͑3͒ are the
lost days that the contractor notified to the owner. Three alleged
lost days January 13, February 6 and 7, 2006 were not agreed to
by the owner ͓noted as “dispute” in Column ͑12͔͒. An independent review of the contemporaneous project records was required
to understand what happened on those days. January 13 appeared
to be an actual lost day because project photos taken on the same
day show muddy site conditions that did not allow concrete trucks
to access the site to deliver concrete for the tilt-up panels.
In contrast, February 6 and 7 were likely not lost days due to
mud. A subcontractor’s daily report shows that the concrete subcontractor could not mobilize in a timely fashion and deliver a
crane for erecting tilt-up panels. The contractor, not the weather,
was responsible for these two lost days. The critical activity ͑erect
tilt-up panels͒ could not be performed because a crane was not

available. Analysis of concurrent delays was not necessary because weather was not a cause on these lost work days. The
independent review also revealed that December 26, 2005 ͑Monday͒ was not a lost day due to weather although heavy rain occurred that day. Daily logs revealed that Christmas Day occurred
on Sunday, December 25, 2005 and was moved to Monday, December 26, 2005. The contractor assigned that day as a holiday in
the project schedule update. The result is that only 37 lost days
were actually caused by weather-related conditions and/or their
lingering effects. The key task was to determine which of those
days were abnormal.
The remaining columns in Table 2 have the following meanings. Column ͑4͒ describes the type of lost day that the contractor
documented in its notices. The two types were “rain” and mud ͑or
dry-out͒. Precipitation amounts ͓Column ͑5͔͒ were obtained from
National Oceanic and Atmospheric Administration ͑NOAA͒
͑2008b͒. Column ͑6͒ shows the equivalent precipitation in millimeters. As previously mentioned, the contract did not provide any
threshold for rains. The lost days recorded mostly depended upon
the actual conditions and contractor judgment. Several rain days
had low precipitation but caused lost time; for example, January
27 ͑0.25 mm͒, February 1 ͑0.51 mm͒, and February 2 ͑1.27 mm͒
of 2006. In contrast, the contractor still worked on layout, forms,
and rebar of tilt panels on days with similar or higher precipitation during work hours; e.g., March 24 ͑0.76 mm͒ and March 28
͑1.78 mm͒ of 2006 ͑not shown in Table 2 because these 2 days
were not notified as lost days͒. One of the reasons was different
construction activities were performed. An arbitrary threshold for
defining adverse weather for the entire project would not work in
this case. A better method would be to determine different thresholds for different activities.
Given the relationship between precipitation and lost days by
rain in this case, Columns ͑7͒ and ͑8͒ count the chronological
order for a particular rain day for a given month for the selected
thresholds: 0.254 mm ͑0.01 in.͒ and 2.54 mm ͑0.1 in.͒, respectively. For instance, the rain on December 21, 2005 was the ninth
͑fifth͒ rain day with a precipitation of 0.254 ͑2.54͒ mm or greater
for that month. Column ͑9͒ shows if an adverse weather day was
abnormal. This column is based on the comparison between the

order of a certain rain day and the monthly anticipated ͑average͒
number of rain days presented in Figs. 2 and 3. For example,
using the threshold of 0.254 mm, the rain on December 21, 2005
was considered abnormal ͑“yes”͒ because it was the ninth rain
day in December of 2005 while 8 days of rain were anticipated

JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / DECEMBER 2010 / 1263

Downloaded 18 May 2012 to 130.220.71.20. Redistribution subject to ASCE license or copyright. Visit


Table 2. Analysis of Adverse Weather for Project ABC, Livermore, California
Precipitation

Number
͑1͒

Notified lost day
͑2͒

͑3͒

Type of
lost day
͑4͒

a

in.
͑5͒

mm
͑6͒

ith rain dayb
Ն0.254

Ն2.54

mm
͑7͒

mm
͑8͒

Abnormal
͑9͒

Critical
͑10͒

Excusable
͑11͒

Remark
͑12͒

n/a ͑n/a͒
No ͑no͒
1

Friday
November 4, 2005
Rain
0.08
2.03
1
1
No ͑no͒d
2
Tuesday
November 29, 2005
Rain
0.48
12.19
4
1
No ͑no͒
n/a ͑n/a͒
No ͑no͒
3
Thursday
December 1, 2005
Rain
0.27
6.86
1
1
No ͑no͒
n/a ͑n/a͒
No ͑no͒

4
Friday
December 2, 2005
Mud
0.20
5.08
2
2
No ͑no͒
n/a ͑n/a͒
No ͑no͒
5
Monday
December 19, 2005
Mud
0.05
1.27
7
n/a
No ͑no͒
n/a ͑n/a͒
No ͑no͒
6
Tuesday
December 20, 2005
Mud
0.01
0.25
8
n/a

No ͑no͒
n/a ͑n/a͒
No ͑no͒
7
Wednesday
December 21, 2005
Rain
0.27
6.86
9
5
Yes ͑no͒
Yes ͑n/a͒
yes ͑no͒
8
Thursday
December 22, 2005
Rain
0.38
9.65
10
6
Yes ͑yes͒
Yes ͑yes͒
Yes ͑yes͒
n/a
Yes ͑yes͒
Yes ͑yes͒
Yes ͑yes͒
9

Friday
December 23, 2005
Mud
0.00
0.00
n/ac
10
Monday
December 26, 2005
Rain
0.37
9.40
12
8
n/a ͑n/a͒
n/a ͑n/a͒
n/a ͑n/a͒
Holiday
11
Wednesday
December 28, 2005
Rain
0.23
5.84
13
9
Yes ͑yes͒
Yes ͑yes͒
Yes ͑yes͒
12

Tuesday
January 3, 2006
Rain
0.02
0.51
2
n/a
No ͑no͒
n/a ͑n/a͒
No ͑no͒
13
Wednesday
January 4, 2006
Mud
0.00
0.00
n/a
n/a
No ͑no͒
n/a ͑n/a͒
No ͑no͒
14
Thursday
January 5, 2006
Mud
0.00
0.00
n/a
n/a
No ͑no͒

n/a ͑n/a͒
No ͑no͒
15
Friday
January 13, 2006
Mud
0.00
0.00
n/a
n/a
No ͑no͒
n/a ͑n/a͒
No ͑no͒
Dispute
16
Friday
January 27, 2006
Rain
0.01
0.25
8
n/a
No ͑no͒
n/a ͑n/a͒
No ͑no͒
Yes ͑n/a͒
yes ͑no͒
17
Monday
January 30, 2006

Rain
0.16
4.06
11
5
Yes ͑no͒
18
Tuesday
January 31, 2006
Mud
0.00
0.00
n/a
n/a
Yes ͑no͒
Yes ͑n/a͒
yes ͑no͒
19
Wednesday
February 1, 2006
Rain
0.02
0.51
1
n/a
No ͑no͒
n/a ͑n/a͒
No ͑no͒
20
Thursday

February 2, 2006
Rain
0.05
1.27
2
n/a
No ͑no͒
n/a ͑n/a͒
No ͑no͒
21
Friday
February 3, 2006
Mud
0.00
0.00
n/a
n/a
No ͑no͒
n/a ͑n/a͒
No ͑no͒
22
Monday
February 6, 2006
Mud
0.00
0.00
n/a
n/a
No ͑no͒
n/a ͑n/a͒

No ͑no͒
Dispute
23
Tuesday
February 7, 2006
Mud
0.00
0.00
n/a
n/a
No ͑no͒
n/a ͑n/a͒
No ͑no͒
Dispute
24
Monday
February 27, 2006
Rain
0.21
5.33
8
4
No ͑no͒
n/a ͑n/a͒
No ͑no͒
No ͑no͒
25
Tuesday
February 28, 2006
Rain

0.44
11.18
9
5
No ͑no͒
n/a ͑n/a͒
26
Wednesday
March 1, 2006
Mud
0.00
0.00
n/a
n/a
No ͑no͒
n/a ͑n/a͒
No ͑no͒
27
Thursday
March 2, 2006
Rain
0.19
4.83
1
1
No ͑no͒
n/a ͑n/a͒
No ͑no͒
28
Friday

March 3, 2006
Rain
0.24
6.10
2
2
No ͑no͒
n/a ͑n/a͒
No ͑no͒
29
Monday
March 6, 2006
Rain
0.54
13.72
3
3
No ͑no͒
n/a ͑n/a͒
No ͑no͒
30
Tuesday
March 7, 2006
Rain
0.12
3.05
4
4
No ͑no͒
n/a ͑n/a͒

No ͑no͒
31
Tuesday
March 14, 2006
Rain
0.31
7.87
9
8
No ͑yes͒
n/a ͑yes͒
No ͑yes͒
32
Friday
March 17, 2006
Rain
0.64
16.26
10
9
No ͑yes͒
n/a ͑yes͒
No ͑yes͒
33
Wednesday
March 29, 2006
Rain
0.30
7.62
17

14
Yes ͑yes͒
Yes ͑yes͒
Yes ͑yes͒
34
Monday
April 3, 2006
Rain
0.37
9.40
3
1
No ͑no͒
n/a ͑n/a͒
No ͑no͒
35
Tuesday
April 4, 2006
Rain
0.95
24.13
4
2
No ͑no͒
n/a ͑n/a͒
No ͑no͒
36
Wednesday
April 5, 2006
Rain

0.21
5.33
5
3
No ͑no͒
n/a ͑n/a͒
No ͑no͒
37
Wednesday
April 12, 2006
Rain
0.31
7.87
10
5
Yes ͑yes͒
Yes ͑yes͒
Yes ͑yes͒
38
Thursday
April 13, 2006
Mud
0.01
0.25
11
n/a
Yes ͑yes͒
Yes ͑yes͒
Yes ͑yes͒
39

Monday
April 17, 2006
Rain
0.07
1.78
14
n/a
Yes ͑no͒
Yes ͑n/a͒
Yes ͑no͒
40
Tuesday
April 18, 2006
Mud
0.00
0.00
n/a
n/a
Yes ͑no͒
Yes ͑n/a͒
Yes ͑no͒
Note: n / a = not applicable.
a
Precipitation obtained from National Oceanic and Atmospheric Administration ͑NOAA͒ ͑2008b͒.
b
The order of a rain day in a particular month when rain is defined by precipitation.
c
n/a in Columns ͑7͒ and ͑8͒, Columns ͑9͒ and ͑11͒, and Column ͑10͒ stands for “no rain” by definition, “not a lost day,” and “not check” if not abnormal,
respectively.
d

In Columns ͑9͒–͑11͒, x͑y͒ is the corresponding outcome per Column ͑7͒ ͓Column ͑8͔͒.

for that month using the same threshold ͑Fig. 2͒. However, that
rain day would not be abnormal ͑“no”͒ if the threshold of 2.54
mm was used.
This case study also shows lingering days or mud days which
the contractor could not perform their critical activities. Since
mud days do not have precipitation, the analysis used above does

not apply. Instead, this analysis adopts the method proposed by
Tennessee Department of Finance and Administration ͑TennDFA͒
͑2007͒ as mentioned previously. This approach states that mud
days caused by “normal” rains are considered normal. Using the
threshold of 0.254 mm, the mud day of January 31, 2006 would
be abnormal because the mud was caused by the abnormal rains

1264 / JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / DECEMBER 2010

Downloaded 18 May 2012 to 130.220.71.20. Redistribution subject to ASCE license or copyright. Visit


2+3#*-# '#(%)#*
(.(",-/!""#$% '#(%)#* +#"(, +(,.!%/0/#+ 1, $!.%*($%!*

4#*# %)# '#(%)#* +#"(,
+(,- ($$#5%#+ 1, !'.#*6
8#%#*9/.# %)# (3#*(>#
.:91#* !0 (+3#*-#
'#(%)#* +(,- 0*!9;
< !.%*($% +!$:9#.%< @/-%!*/$(" '#(%)#* +(%(

< A)*#-)!"+ 3(":#-

E!
8#%#*9/.# /0 +/-5:%#+ +(,- '#*#
*#("", +#"(,- 0*!9;
< 5*!=#$% *#$!*+< %/9/.> !0 (+3#*-# '#(%)#*
< $*/%#*/( 0!* "!-% +(,< %,5#- !0 '!*?
< "/.>#*/.> #00#$%-

F#-

7/-% '#(%)#* +#"(, +(,("$:"(%# %)# .:91#* !0
(1.!*9(" '#(%)#* +#"(, +(,)#$? /0 (1.!*9(" '#(%)#*
+#"(, +(,- '#*# $*/%/$(";
< 5*!=#$% -$)#+:"#< %,5#- !0 '!*?
!.3#*% %)# .:91#* !0
(1.!*9(" (.+ $*/%/$(" "!-% +(,/.%! $("#.+(* +(,- 1,;
< -$)#+:"/.> -!0%'(*#B !*
< -#(-!.(" '#/>)%/.> !0 +(,CD%#.-/!.- !0 %/9# +:# %!
:.0!*#-##. '#(%)#* $!.+/%/!.-

Fig. 5. Flowchart of adverse weather analysis

on January 30, 2006 and earlier days. Conversely, the same day
would be normal with a threshold of 2.54 mm. Column ͑10͒
shows whether the work impacted by the weather conditions was
critical based on the delay analysis of contemporaneous project
schedule updates. Detailed analysis is beyond the scope of this
paper.
Column ͑11͒ shows if a qualified lost day is excusable ͑yes͒ or

inexcusable ͑no͒. A qualified lost day is excusable only if it was
abnormal and critical. In other words, yes must appear on both
Columns ͑9͒ and ͑10͒.
Time Extension
The owner should grant a time extension in calendar days for
unusually severe weather to the contractor in this contract. For the
precipitation threshold of 0.254 mm, the number of excusable

work days lost due to abnormal weather conditions would be 11.
That number would be eight work days if the threshold of 2.54
mm is used ͑Table 2͒.
These work days lost due to abnormal weather need to be
converted into calendar days. As previously discussed, three alternatives for the conversion include the fixed conversion factor
͑7/5͒, the seasonal weighting of days, and the schedule fragnet for
unusually severe weather. While this last alternative is the most
reasonable, it cannot be applied in this case because the project
schedules were not assigned all anticipated nonworking days
͑U.S. holidays, union holidays, anticipated weather days͒. Using
the conversion factor of 7/5, the time extension would be approximately 15 calendar days for the precipitation threshold of 0.254
mm and 11 calendar days for the precipitation threshold of 2.54
mm ͓hereafter 15 ͑11͒ calendar days for the precipitation threshold of 0.254 ͑2.54͒ mm͔.
Table 3 presents the use of the seasonal weighting of days for
the conversion. The calculation of available work days is similar
to the above example. Under this alternative the extension of time
would be 23 ͑14͒ calendar days for the precipitation threshold
value of 0.254 ͑2.54͒ mm ͑Table 3͒. The number of excusable lost
days ͑expressed in work days͒ for a particular month in Table 3
equals to the number of yes in Column 11 of Table 2 for the
corresponding month. In March 2006, for instance, the numbers
of excusable lost days are one and three based on the threshold of

0.254 and 2.54 mm, respectively ͑Table 2͒. The difference in time
extension for the two alternatives is eight ͑three͒ calendar days
for the different weather thresholds. The difference in liquidated
damages assessed is $40,000 ͑$15,000͒, which can be a significant amount for a $5 million contract.

Discussion
The case study draws many interesting points with reference to
factors affecting the analysis of severe weather. A consistent
analysis of severe weather is impossible if the contract does not
define what constitutes normally severe weather and determines
weather thresholds for different work to be performed. Such definitions are not always available, and as this case study shows,
making different assumptions for such absent definitions results in
different time extensions. For instance, by adopting Tennessee
Department of Finance and Administration ͑TennDFA͒ ͑2007͒ the
above analysis presumes that mud days caused by normally severe weather are normal. The analysis cannot be done properly
without that presumption, because the anticipated severe weather
delay days including mud days cannot be established. However,

Table 3. Calendar Days Lost Using the Seasonal Weighting of Days
Available work daysa
Month

Ն0.254

November 2005
December 2005
January 2006
February 2006
March 2006
April 2006

Winter 2005/2006
a
Available work days, lost

15
15
13
13
16
16

Excusable lost daysa ͑work day͒

Ն2.54

Calendar days

18
17
16
15
19
18

30
31
31
28
31
30

Ն0.254

Ն2.54

0
0
4
3
2
0
0
0
1
3
4
2
11
8
days ͑work or calendar day͒ based on thresholds ͑0.254 mm or 2.54 mm͒.

Excusable lost daysa ͑calendar day͒
Ն0.254

Ն2.54

0
8
5
0

2
8
23

0
5
—
0
5
3
14

JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / DECEMBER 2010 / 1265

Downloaded 18 May 2012 to 130.220.71.20. Redistribution subject to ASCE license or copyright. Visit


lingering days that roll over to a following month are still challengeable. The mud day on March 1, 2006 was caused by consecutive rain days at the end of February. The fact that whether
that mud day would have been normal or abnormal if rains on
February 27 and 28 had been abnormal is uncertain. Fortunately,
the above analysis does not encounter this dilemma because the
rains at the end of February 2006 were normal ͑Table 2͒.
Another conclusion is that contracts should specify monthly
anticipated severe weather delay days. Contractors should account for those days when planning their work and preparing their
baseline schedules. At best, these delay days should include anticipated lingering days based on construction activities scheduled
in the periods of the anticipated severe weather conditions. In the
case that only normally adverse weather days provided, the contract should determine whether a lingering day rolling over to the
following month is normal or abnormal given that adverse
weather causing that lingering day is abnormal or not.
Different assumptions on the threshold values of precipitation

change the results of the analysis. With two thresholds used in
this case study, the results are substantially different. A particular
rain day can be normal for a given threshold but abnormal if
another threshold is used. The results change considerably if the
comparison for usually versus unusually severe weather were
based on the whole project duration or the whole period allegedly
affected by weather rather than a month-by-month basis. That is,
the winter of 2005/2006 in the case study. In addition, a predetermined threshold for the entire project may not totally decide if a
particular weather condition causes lost day. With the monthly
anticipated adverse weather days specified in contracts, thresholds
should not be the primary factor deciding if a day is lost. Any
adverse weather day and its associated dry-out day͑s͒ ideally will
be agreed to by the involved parties. If not, the threshold together
with the type of work scheduled, work quality required, the portion of work performed, site conditions, and so on will decide lost
days.
In any circumstance, threshold values for different weather
conditions ͑e.g., rain/snow, wind, temperature͒ should be defined
in contracts, ideally for different work, activities or trades. They
are not only a source of reference for determining lost days, if the
parties do not agree as mentioned, but also are a baseline for
counting actual adverse days in a particular month ͓Columns ͑7͒
and ͑8͒ in Table 2͔. Consequently, whether a lost day is abnormal
is fairly determined.
Conversion from work days lost to calendar days lost is another issue to be addressed in contracts. We believe the conversion method proposed by U.S. Army Corps of Engineers
͑USACE͒ ͑2008͒ should be adopted. That is, time extensions for
unusually severe weather should be determined based on schedule
impact analysis. However, the seasonal weighting of days can be
used with similar results. The use of the 7/5 conversion factor that
is widely employed in the industry may not be equitable. Therefore, depending on project scale and the sophistication of the
parties involved, contracts should specify a conversion option,

either adopting U.S. Army Corps of Engineers ͑USACE͒ ͑2008͒
or the seasonal weighting of days to avoid confusion and speculation.
The weather analysis in the case study does not cover such
factors as the type of work, criteria for lost days, and lost days
equivalent due to lost labor productivity. In this case, the data
were not available for further analysis. However, these factors
very often affect the schedule analysis of unusually severe
weather. Contemporaneous schedule updates show that many
critical activities were significantly delayed during the 2005/2006

winter. The delay was beyond the direct lost days due to severe
weather.
In this particular case, factors such as site access, logistics,
interrupted learning curve, remobilization of manpower and
equipment, and weather change may have negatively impacted
labor productivity. That lost productivity could in turn significantly delay this project but the data to analyze such were not
available. Consequently, a truly reasonable time extension might
be greater than the time extensions computed from the methods
outlined in this paper. Contracts should provide a mechanism for
dealing with such lost days equivalent.
As to the reality of this particular case, the general contractor
demanded that all of 40 weather delay days notified were excusable delays. The owner on the other hand accepted only 21 calendar days without any justification. The two parties never agreed
each other on weather based time extensions, so they took this
issue to binding arbitration. In his decision the arbitrator only
ruled on merit and damages without actually ruling on how many
delay days were attributable to unforeseen weather conditions.
Finally, though the case study is a commercial building project,
analysis of adverse weather presented in this paper can be used
for other types of projects, for example, infrastructure projects.

Conclusions
Acceptable time extensions for unusually severe weather can be
difficult to establish. This is a source of costly claims and disputes
in construction. This research has discussed some key factors that
significantly affect the analysis of severe weather for time extensions. They include the definition of normal weather, weather
thresholds, the type of work, lingering days, criteria for lost days,
lost days equivalent due to lost productivity, and the conversion
from work days to calendar days lost.
This paper also presents an actual weather-caused delay claim
to demonstrate the sensitivity and significance of those factors.
This case study demonstrates that the delay analysis results can be
substantially different, depending on different interpretations of
those factors in weather analysis. In this actual case the contractor
claimed 40 work days as excusable while the owner accepted
only 21 calendar days. The analysis provides 23 and 14 calendar
days with precipitation thresholds of 0.254 and 2.54 mm, respectively.
The paper also illustrates that unequivocal contract terms can
play a critical role in reducing such discrepancies. Monthly anticipated weather delay days, along with their source, should be
included in the contracts. Also, a method to determine lingering
days and the definition of anticipated lingering days should be
clearly defined in contracts. Threshold values for weather parameters should be specified in the contracts. Depending on the
project, weather thresholds should be provided for either the entire project or different work activities/trades. Weather thresholds
play a critical role when differentiating unusually severe weather
from usually severe weather. Contracts also need to define a
method to convert a time extension from work days to calendar
days. Ideally, any time extension should depend on the contemporaneous time impact of the lost days caused by unusually severe weather. In this case the project’s calendar must clearly and
adequately define weekends, holidays and monthly anticipated
weather delay days. Otherwise, the contracts should provide, or at
least describe, procedures to develop a seasonal weighting of
days.

Finally, contracts should address the requisites to request a

1266 / JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / DECEMBER 2010

Downloaded 18 May 2012 to 130.220.71.20. Redistribution subject to ASCE license or copyright. Visit


time extension caused by inefficiency due to unusually severe
weather conditions. Future research may provide a proper method
for analyzing lost days equivalent as a result of lost productivity
due to adverse weather. Analysis presented in this paper does not
quantify the impacts of lost productivity caused by severe weather
on the allowable number of lost days.

References
American Institute of Architects. ͑1997͒. “General conditions of the contract for construction.” Document A201-1997, Washington, D.C.
American Institute of Architects. ͑2007͒. “General conditions of the contract for construction.” Document A201-2007, Washington, D.C.
Attanasi, E. D., Johnson, S. R., LeDuc, S., and McQuigg, J. D. ͑1973͒.
“Forecasting work conditions for road construction activities: An application of alternative probability models.” Mon. Weather Rev.,
101͑3͒, 223–230.
Baldwin, J. R., Manthei, J. M., Rothbart, H., and Harris, R. B. ͑1971͒.
“Causes of delay in the construction industry.” J. Constr. Div., 97͑2͒,
177–187.
Benjamin, N. B. H., and Greenwald, T. W. ͑1973͒. “Simulating effects of
weather on construction.” J. Constr. Div., 99͑1͒, 175–190.
Bramble, B. B., and Callahan, M. T. ͑2000͒. Construction delay claims,
3rd Ed., Aspen Law & Business, New York.
Chan, E. H. W., and Au, M. C. Y. ͑2008͒. “Relationship between organizational sizes and contractors’ risk pricing behaviors for weather risk
under different project values and durations.” J. Constr. Eng. Manage., 134͑9͒, 673–680.
Connors, R. B. ͑2003͒. “Weather derivatives allow construction to hedge

weather risk.” Coastal Eng., 45͑3͒, 21–24.
El-Rayes, K., and Moselhi, O. ͑2001͒. “Impact of rainfall on the productivity of highway construction.” J. Constr. Eng. Manage., 127͑2͒,
125–131.
El-Razek, M. E. A., Bassioni, H. A., and Mobarak, A. M. ͑2008͒. “Causes
of delay in building construction projects in Egypt.” J. Constr. Eng.
Manage., 134͑11͒, 831–841.
Finke, M. R. ͑1990͒. “Weather-related delays on government contracts.”
Proc., AACE Int. Transactions of the Annual Meetings, The Association for the Advancement of Cost Engineering International ͑AACEI͒,
Morgantown, W.Va.
Grimm, C. T., and Wagner, N. K. ͑1974͒. “Weather effects on mason
productivity.” J. Constr. Div., 100͑3͒, 319–335.
Handex of the Carolinas, Inc. v. County of Haywood, and Municipal
Engineering Services Company, P.A., No. COA03-1658 ͑Court of Appeals of North Carolina 2005͒.
Hegazy, T., and Menesi, W. ͑2008͒. “Delay analysis under multiple baseline updates.” J. Constr. Eng. Manage., 134͑8͒, 575–662.
Hinze, J., and Couey, J. ͑1989͒. “Weather in construction contracts.” J.
Constr. Eng. Manage., 115͑2͒, 270–283.
Kartam, S. ͑1999͒. “Generic methodology for analyzing delay claims.” J.
Constr. Eng. Manage., 125͑6͒, 409–419.
Kenner, S., Johnson, R. L., Miller, J. R., Salmen, J. A., and Matt, S. A.
͑1998͒. “Development of working day weather charts for transporta-

tion construction in South Dakota, study SD97-07.” Technical Rep.,
South Dakota Dept. of Transportation, Pierre, S.D.
Koehn, E., and Brown, G. ͑1985͒. “Climatic effects on construction.” J.
Constr. Eng. Manage., 111͑2͒, 129–137.
Laufer, A., and Cohenca, D. ͑1990͒. “Factors affecting construction planning outcomes.” J. Constr. Eng. Manage., 116͑1͒, 135–156.
Maunder, W. J., Johnson, S. R., and McQuigg, J. D. ͑1971͒. “Study of the
effect of weather on road construction: A simulation model.” Mon.
Weather Rev., 99͑12͒, 939–945.
McDevitt & Street Co. v. Marriott Corp., 713 F. Supp. 906 ͓U.S. District

Court Eastern District of Virginia ͑E.D. Va.͒ 1989͔.
McDonald, D. F. ͑2000͒. “Weather delays and impacts.” Cost Eng.,
42͑5͒, 34–39.
Moselhi, O., and El-Rayes, K. ͑2002͒. “Analyzing weather-related construction claims.” Coastal Eng., 44͑8͒, 12–19.
Moselhi, O., Gong, D., and El-Rayes, K. ͑1997͒. “Estimating weather
impact on the duration of construction activities.” Can. J. Civ. Eng.,
24͑3͒, 359–366.
National Oceanic and Atmospheric Administration ͑NOAA͒. ͑2008a͒.
Climatography of the United States no. 20 1971-2000: Livermore, CA
station, Asheville, N.C., ͗ />clim20/ca/044997.pdf͘ ͑Nov. 8, 2008͒.
National Oceanic and Atmospheric Administration ͑NOAA͒. ͑2008b͒.
Record of river and climatological observations: Livermore, CA station, Asheville,
N.C.,
͗ />coop.html͘ ͑Nov. 8, 2008͒.
Nguyen, L. D., and Ibbs, W. ͑2008͒. “FLORA: New forensic schedule
analysis technique.” J. Constr. Eng. Manage., 134͑7͒, 483–491.
Semple, C., Harman, F. T., and Jergeas, G. ͑1994͒. “Construction claims
and disputes: Causes and cost/time overruns.” J. Constr. Eng. Manage., 120͑4͒, 785–795.
Smith, G. R., and Hancher, D. E. ͑1989͒. “Estimating precipitation impacts for scheduling.” J. Constr. Eng. Manage., 115͑4͒, 552–566.
Tennessee Department of Finance and Administration ͑TennDFA͒.
͑2007͒. “Section 01 26 20—Weather delays.” Designer’s manual,
Nashville, Tenn.
Thomas, H. R., Riley, D. R., and Sanvido, V. E. ͑1999͒. “Loss of labor
productivity due to delivery methods and weather.” J. Constr. Eng.
Manage., 125͑1͒, 39–46.
U.S. Army Corps of Engineers ͑USACE͒. ͑1989͒. “Construction time
extensions for weather.” Regulation No. ER 415-1-15, Vicksburg,
Miss.
U.S. Army Corps of Engineers ͑USACE͒. ͑1996͒. “Developing and analyzing adverse weather data.” Construction Bulletin No. 96-13, Vicksburg, Miss.
U.S. Army Corps of Engineers ͑USACE͒. ͑2008͒. “Developing, analyzing

and implementing adverse weather policy for construction contracts.”
Engineering and Construction Bulletin No. 2008-23, Vicksburg, Miss.
Worthington, M. M., and Goldsman, L. P. ͑1998͒. Contracting with the
federal government, 4th Ed., Wiley, New York.
Xi, Y., Balaji, R., and Molenaar, K. ͑2005͒. Quantifying construction
delays due to weather, Federal Highway Administration, Central Federal Lands Highway Div., Lakewood, Colo.
Yogeswaran, K., Kumaraswamy, M. M., and Miller, D. R. A. ͑1998͒.
“Claims for extensions of time in civil engineering projects.” Constr.
Manage. Econom., 16͑3͒, 283–293.

JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / DECEMBER 2010 / 1267

Downloaded 18 May 2012 to 130.220.71.20. Redistribution subject to ASCE license or copyright. Visit