Unless you are working indoors, weather is always a consideration when building a CPM schedule. Somehow, normal weather must be addressed for any work that can be impacted by inclement weather. Our only concern should be normal weather; unusual weather is an excusable delay. This of course raises the issue of how do we determine what exactly is normal weather? Contracts often mention that time extensions will only be granted for abnormal weather without defining normal weather. It is easy to find weather data: in the United States the best source of historical weather conditions would be the National Oceanic and Atmospheric Administration. The NOAA has records for thousands of weather stations around the country that in some cases go back a hundred years or more.
Even so, there is not a single standard for applying NOAA data. Should we take the average for the past four years to determine what is normal? Five years? Six? Federal contracts generally rely upon an average of the last ten years to determine normal weather conditions. (As a personal aside, I have lived in California for 11 years and the weather during the past four years has been radically different than what I first encountered in 2004). Most private construction contracts are unfortunately silent as to what sort of average might be considered reasonable.
The U.S. Army Corps of Engineers is probably the best example of how to specify normal weather. The USACE typically tells contractors how many days of inclement weather to include in the CPM schedule each month. NOAA data would only tell us the average temperature and precipitation, which leaves open to interpretation how a day with, say, 0.1″ of precipitation should be treated. For good or bad, the USACE specifications leave no doubt how many days should be blocked out for weather – not including weekends and holidays. Contractors can not claim they thought it would only rain or snow on weekends!
Some State agencies use a methodology similar to the USACE. Last year I was teaching a highway contractor in Minnesota how to schedule projects using Primavera P6. Someone pulled out the specifications for an upcoming project. Glancing down to the weather provisions, I was not surprised to see that the Minnesota Department of Transportation expected contractors to block out 20 days for inclement weather in January beyond weekends and holidays. In case you are wondering, that wipes out the entire month! I was there in March and it was still below zero degrees in the morning. Winter work is nearly impossible outside unless you are ice fishing.
Sometimes, the contractor does not have to address inclement weather at all. The California Department of Transportation (CalTRANS) specifies contract durations in working days. As the project moves along CalTRANS tallies only the days the contractor is able to works. Bad weather days are not counted. As you might have guessed, this means the project end date shown in the baseline schedule assumes perfect weather and therefore is unlikely to be maintained.
Once we have established some sort of standard for normal weather, we can then move on to a strategy for incorporating this weather into the CPM schedule. Below are the three basic strategies that I use, in my order of preference:
I. Add Normal Weather to the Work Calendar
If the owner has already told me how many days of normal inclement weather to anticipate, it is logical to block out this number of days as if they were holidays. While it is not possible to distinguish a weather day from a holiday in Primavera P6, I try to put my weather days on any Tuesdays, Wednesdays and Thursdays to avoid Monday holidays like Labor Day, Memorial Day and Presidents Day. Obviously Thanksgiving is always on a Thursday but otherwise this works pretty well. NOAA data can even tell us which days of the month typically have the most precipitation if we want to find the best candidates for weather days. I also have a Project Notebook Topic I created in Primavera P6 called “Weather” that I use to list the weather days included in the schedule. This removes any doubt as to why these days were blocked out.
This approach does require two work calendars, however. One work calendar will be for the weather-sensitive work while the other will not include any weather days. After all, shop drawings are not affected by weather and there are usually some field activities that take place indoors. Otherwise the two calendars will probably share the same holidays.
At the end of the month, many schedulers like to replace the planned weather days with the actual weather days. This creates a historical record of when the inclement weather occurred. But actual weather days are presumably being tracked elsewhere so I consider this step to be optional.
II. Create a Contingency Activity for Normal Weather
I started using this strategy in the 1980s as a way to show early completion. The contractor would plan to finish the project early so we needed an activity to bridge the gap between the early completion milestone and the contractual completion milestone. Not all owners would accept the contractor’s right to finish early. In some cases the owner would issue a no-cost change order to modify the contract completion date. Basically, the owner was calling the contractor’s bluff. If the contractor figured he could finish early then why not make that the new contract completion date? Otherwise, the contractor might submit a delay claim based on not being able to finish the project early even though the owner never requested the earlier completion date.
On some projects the bid documents are held in escrow; if a delay claim is submitted the bid documents can be reviewed to see if the contractor based his overhead costs on the shorter project duration. A CPM schedule that shows early completion is often viewed with suspicion unless there is further proof. But in some cases the contract documents specify that early completion is not allowed. The contract effectively becomes a professional services agreement with a fixed time frame.
Normal inclement weather can also be treated as a contingency. The number of expected weather days are added up and inserted into an activity between early completion and contract completion. In this case, however, we are not really expecting to finish early; the contingency will disappear if the total amount of normal inclement weather is realized. During each update we reduce the remaining duration of the contingency activity by the number of actual weather days incurred. In theory, there will be no more weather days once the remaining duration reaches zero days. Otherwise, the contractor is entitled to a time extension.
Both the first and second strategies require an analysis to determine how many weather days are to be expected over the life of the project. Unless the number of weather days are specified in the contract there could be disagreements as to how many days should be included. A smaller number helps the contractor with delay claims while a bigger number protects the owner against the very same claims. Activity durations should be based on perfect weather, as normal weather is accounted for by either the calendar or the contingency.
III. Add Normal Weather to the Activity Durations
This is the oldest strategy and does not require as much effort as the other strategies. The contractor simply adds additional time to the activity durations based on the expected weather. I list this strategy last because it is nearly impossible to verify how much time was added for weather unless the contractor keeps detailed notes (such as using an Activity Notebook Topic in Primavera P6 to explain where the days were added). Moreover, the contractor needs to know what time of year each activity will take place before he can add the right amount of time. For this reason schedule logic is needed first so that the start and finish dates are reasonably accurate. But the reason I list this strategy last is that if an activity is delayed the added time for weather may now be too much or too little. Constant monitoring is required to ensure that activities have not moved into a time frame with different weather conditions.
Update: a member of the Association for the Advancement of Cost Engineering took exception to this third method after my original post was published. He felt this method should never be used and referred to the AACE’s Recommended Practice for addressing weather in a schedule. But he was missing the point. I have used this third method for years without any issues. Keep in mind that if the weather days are shown on the calendar or as a contingency activity the contractor has boxed himself into a corner. He cannot later claim he expected fewer normal weather days.
The contractor may very well want to give himself some “wiggle room” to clarify his understanding of normal weather once a dispute arises. And who can blame him if the owner does not specify the number of weather days either? The owner is legally responsible for any ambiguities in the contract documents. Conversely, if the contractor feels he has been treated fairly by the owner on all other matters, he may be willing to concede a few weather days that he otherwise felt entitled to claim.
Let’s use liquidated damages as an analogy. Liquidated damages clauses have been around for decades. A contractor is put on notice that a specific amount of money will be withheld by the owner for each day he is late. Liquidated damages are an approximation of the financial impact caused by the project being delivered late. As such, liquidated damages cannot (and are not required to) be proven. The contractor may think the liquidated damages are outrageous, but this is not a matter for negotiation.
Yet the same owner has no clue how many days should be included in the schedule for normal inclement weather? There are multiple sources of historical data to help him make this decision. So perhaps the owner wants a little “wiggle room” as well. He has the information necessary to make a decision; he just decided to leave the contract ambiguous. Shame on him.
The bottom line is that most construction contracts do not adequately address the definition of normal weather. AACE’s Recommended Practice is not the solution either since it still leaves this matter open to interpretation.
What is your favorite method of addressing inclement weather? I can be reached here.
