(1) Explain your problem, don't simply post "This isn't working". What were you doing when you faced the problem? What have you tried to resolve - did you look for a solution using "Search" ? Has it happened just once or several times?
(2) It's also good to get feedback when a solution is found, return to the original post to explain how it was resolved so that more people can also use the results.
Dear All Planners,
Which the best Monte Carlo computer program I can use as Scheduler and Planner to estimate the risk in my project?
Thank you,
HANSON
Printer-friendly version
Send to friend
Dennis,
Sorry I mixed the name, I was referring to you. Tomorrow I will e-mail you whatever I have or perhaps will post it for download by anyone interested to join the discussion. It seems the data is not that difficult to get after I got an excellent and fast response from NOAA.
Will use Excel 97-2003 xls file format if this is ok.
Keep in touch.
Best Regards,
Rafael
Rafael:
You refer to "David", do you mean me? If so, I can send you the data as an .xlsx file or any earlier version of Excel. Threshhold was .1", but that can be adjusted once there is a meaningful threshold established. What is your current experience with number of construction days lost to rain by month? That will go a long way towards establishing the threshold, but even here we may have to complete the analysis using different thresholds.
There are about 330,000 rows of raw data, so do not look for immediate answers unless we get some programming help. Send me an email request and I will attach the file to the reply.
David,
RE: My preliminary analysis for SJIA suggests that it will be more on the order of 8 days per month (8.41) with a range from 4-14 days, but that is only one year (1967) and this would be further reduced if there was a make-up Saturday available, say 4 days per month, so you could expect to lose 4-5 days per month to rain if you were working near the San Juan International Airport. My limited experience to date with the variability of weather data would be to question any analysis done on only one year. Is 4 days per month closer to actual experience? What is actual experience?
If you have during a month a single activity that is impacted by a rain calendar changing the dates but not the amount of rain dais is not going to make any difference, is going to be dead wrong.
Of course first things first checking on the average before going into the distribution is not a bad idea., what you are getting for the average seems reasonable for a single month of a single year. Yes at times this will happen, at time will be no rain, at times will be more.
Wow 7 days is a full week, was it a hurricane month? Seems like a close encounter but not a direct hit, still too much in the absence of a hurricane close encounter. It seems to me like the numbers NOAA yields for a mere 1/8 in of rainfall, by our standards almost no rain at all if on a single day.
We have no issues with NOAA other than how difficult it is to get the data, we have issues with the threshold that yield way too many days expected to impact the job making it impossible for the contractor to be able to claim rain days on months out of hurricane season.
The following are NOAA numbers, days with rain bellow .20 or .30 here do not make it enough to stop site works. Seems like you are using 0.10 in as the threshold and it is way too little rain.
Can we move any further discussion to the Risk Forum where we are having a parallel discussion on the theme? I find your questions of most interest and do not want to miss a single one, in addition I want to avoid keeping out of this thread central topic.
Best Regards,
Rafael
Rafael:
My preliminary analysis for SJIA suggests that it will be more on the order of 8 days per month (8.41) with a range from 4-14 days, but that is only one year (1967) and this would be further reduced if there was a make-up Saturday available, say 4 days per month, so you could expect to lose 4-5 days per month to rain if you were working near the San Juan International Airport. My limited experience to date with the variability of weather data would be to question any analysis done on only one year. Is 4 days per month closer to actual experience? What is actual experience?
I thought I provided the link. http://gis.ncdc.noaa.gov/map/viewer/#app=clim
I can provide the Excel workbook, if you would like. I am using Excel 2010, but can save in most any of the earlier versions. This workbook will be available to any who would like to work on this issue with us.
Dennis,
Yes this is what they say and I noticed it on the spot, but if you apply this threshold to most Puerto Rico locations it means it is expected that on average 50% of the days you will go home without performing any work at all. This is far from real, would yield absurd models.
I would like the link for the data you have, for a couple of days I have not being able to find such data. Your contribution is of much value.
Best regards,
Rafael
Rafael:
Thanks for the link. As a spec. for North Carolina an average temp of <40 degress might be applicable, but probably not everywhere: "Normal bad weather days should be those days on which precipitation is 0.10 of an inch, or greater, or any day the temperature fails to exceed an average of 40° F and that bad weather shall have had an adverse effect on the schedule."
BTW: I have the precipitation data by hour for San Juan International Airport that I am now looking at. Probably will not do anything with it, except play a little bit. Date range 01Jan67 to 01Aug12 (~45 years).
Dennis,
Take a look at item 3 in the following reference.
http://www.cagc.org/contractors_bldg/NC_JCR/nc_jcr32_construction_time_for_projects.html
“3. Normal bad weather days should be those days on which precipitation is 0.10 of an inch, or greater, or any day the temperature fails to exceed an average of 40° F and that bad weather shall have had an adverse effect on the schedule. ”
I already posted the issue on the Schedule Risk Forum as you suggested. I am using a specific location so everyone contributing can use the same data. I am still in search for data as I cannot find enough summary details to create a probability distribution that is real, perhaps will have to look at the data for every month of every year out of 30 years and from it generate myself the distribution curve, for this your help will be appreciated.
These are daily running averages, it does not means that on average every day it rained. Somewhere I read the monthly average number of August days with rain over 0.1" is 17 days. I am also looking for what the running average means for NOAA, believe it is a 27 days running average. I am used to the running average concept in the reporting of concrete tests, the American Concrete Institute uses the average of the las 3 consecutive sample 20 days concrete tests as the running average for ultimate strength design wile it was the average of last five for old working stress design method.
It seems like a PHd in hydrology is required to properly interpret the data, I find it abusive to require such burden on the regular contractor for regular jobs. If not, if it is so easy why don't they provide some guidance?
Best regards,
Rafael
Rafael;
Remember these are daily totals (0.1"). If you could point me the the AGC and AIA studies that mention .1", I would appreciate it. Thanks.
Dennis,
Very interersting, weld versus soil, definitivelly different rain calendars for different activities, cannot agree more.
About the 0.3" appropiate for highway construction seems good, althoug not sure, but the 0.1" suggested by the AGC and AIA North Carolina Chapters (less than a mere 1/8") to us is dessert like conditions.
http://www.cagc.org/contractors_bldg/NC_JCR/nc_jcr32_construction_time_for_projects.html
Because I live in Bayamon, closer to Dorado than to San Juan, I will suggest to use Dorado Puerto Rico NOAA and NWS.
To say that on average 11 months out 12 are expected to be lost to rain in Dorado is super-exagerated. It is hard to interpret the data as running average does not means it rains every day. The effect of this is intended to prevent the contractor from claiming rain days. What about September and November? If it comes out this way then it is a dirty trick.
Usually the language of our State and Municipal agencies is precise and clear, they tell you righ way the exact number of expected rain days to be considered contractor risk in an universal languages of numbers, either 1, or 2 or 3 or 4 as I do not recall something over 4 days, 5 is a big number. But they are learning the Federal Government ways of making it hard or impossible for the contractor to present his case.
Best Regards,
Rafael
Rafael;
Did not get much out of either study other than the first study used .3" for its rain days. This may be appropriate for highway construction, but I think weld integrity (porosity) and safety would suggest a more conservative metric - say .1". I used .2" in my analysis, but if I were to do it again, and I expect I will, I am pretty sure I would use .1".
As significant as weather is (IMO) very little constructive guidance is available to predict its impact. Some contracts will specify weather allowances and others will not. Those that do, do not say how they are developed, so any attempt at ranging/risking is so subjective as to be almost meaningless. I welcome your continued insights. Thanks.
FWIW: This is a video of my method. http://www.youtube.com/watch?v=olYF0N2DxCo
I can share some of the excel workbooks used in its development, if you like.
[Note to others: please do not ask for copies of the workbooks unless you have made a substantive contribution to this thread (or the one that may follow]
Dennis,
I will follow your recommendation to move the discussion to the Risk Forum early next week, I will address the issue on how to convert rain data to non work days and will try to expand the procedure to get a statistical distribution instead af a fixed number for expected non work rain days per month.
Before starting the tread will be doing some homework. I found the following references that might be of interest and would welcome you to take a look at them and make the comments you wish to fire the debate.
http://www.sddot.com/business/research/projects/docs/SD1997_07_final_report.pdf
http://ascpro0.ascweb.org/archives/cd/2008/paper/CERT252002008.pdf
We have no problem in identifying days lost to rain after the facts but I am having problem with interpreting how many days are risk of the contractor and how many are a shared risk. I usually find the contract language so poor that it can be interpreted in so many ways I believe any court would accept whatever reasonably was interpreted by the party who did not wrote these vague contract terms. Although vague I am looking for a reasonable methodology that makes sense, if the owner is not clear I still must assume something reasonable.
Best regards,
Rafael
Rafael;
You allude to an interesting dilemma - what is a rain day and how can you forecast its occurrence?
There seems to be no empirical method to determine when work will be stopped for rain. The two methods that have been presented to me are the number of rain drops (18-?) that fall on a brick or circle in a certain time - usually a minute. So we start from here with the 'fact' that we cannot even agree on when to stop work. So you can see how difficult it will be to determine how many rain days we may have in a particular month. Chris's method is as good as anyone's. My approach was to take the NOAA historical rain data back to 1965 (in this case Lake Charles, LA) and then select the number of days in the month when the rain total was more than .2" (this was arbitrary and given the measuring tools previously cited - brick/circle, I would probably use .1" next time)[note:this is not a trivial exercise] then determine the standard deviation and multiply that by a Z value of 1.28 to get a P10/90 value and I now have a min/lik/max value for the month that I can use with OPRA's Weather Module.
Chris and I disagree on his Weather Calendar approach because it does not take into account extreme variability in weather, as you will soon discover when you review the data. That aside, there does not seem to be a 'best' practices here. I too, would welcome the insights of others.
Note: You may receive more comments if this thread were moved to the Risk forum.
Thomas,
Thanks for your response, eventually I will be interested in learning from your approach to weather modeling and how best to determine when rain is to be considered enough to represent a day lost, your contribution will be appreciated.
Usually in our contracts is defined when it is cause to stop work on over 50% of your activities. It might be because it rained last night so much as not to allow you to start work, it might be because is is raining during activities performance.
There is much rain data but it is not clear how to translate this into your model. Stoppage might depend on how saturated is the soil, it is a complex functions of rain intensity, duration, site drainage and soil characteristics among others. Even the things you can do can have an impact on whether you can continue working or not, for a mat foundation you might or might not use a lean or gravel base that will allow you to continue work earlier.
Determining future impact will never be an exact science but we need some guidance to make better models and to justify our assumptions in case of a claim.
What I have seen at the best are attempts but none good enough. The software provides good functionality the problem is on what are the values to be entered.
The following is one of the most interesting articles I have seen, I like the approach but still find it short by a narrow gap.
http://www.slideshare.net/hilalitani/planning-for-adverse-weather-in-construction-projects
Best regards,
Rafael
Rafael - We currently support weather event modeling. We are extending this to support seasonal weather modeling as well as direct ties to web-based databases of hurricane/tidal patterns etc.
Hi,
How can a Risk Analyzer Program identify the Risk in an schedule?
IN my opinion, Risk are those that are possibly encountered during the execution.
Examples are as follows but not limited to:
1. Those mentioned by Rafael, The Weather Condition
2. Strikes by Workers due to unpaid Wages
3. Major Changes introduced by Client that are waived to be claimed
4. Long Lead Items originating from Abroad which can not reach the destination on time.
5. Any Other experiences encountered by EXPERTS.....
If this Program can identify the Risk, why don't we just feed this program with the Project Information and it will create the Schedule so that the Risk will be avoided?
As I understand, Risk are being identified by the person who creates the schedule. The scheduler is supposed to do the "What If" scheduling.
Regards,
Tom,
Weather is one of the major uncertainties at home, weather is a function of calendar not a function of activity duration. If an activity is delayed into rainy season it will impact the activity work calendar but outside rainy season the impact is minimum. Monte Carlo without weather modeling is unrealistic in our case. I suppose in places with very cold winter season snow fall and winter storms are also of concern.
Just to make sure I got it right, does Acumen models weather applying the weather risk to calendars?
What if rain on a single day means work cannot be done on the activity for 2 days until rain effect clears out, how does Acumen Risk deals with this issue on its Monte Carlo runs?
It is not just about some rain occurring in a single day but continuous day rains, perhaps 4 to 10 hours a day, enough to make it a non-work day for each consecutive day plus a recovery period.
And what about inflation, is it a risk event tied to calendars?
Best regards,
Rafael
Hanson
Acumen Risk is worth a look. It is fairly new to the market and therefore does not bring any "baggage" with it. It's a fresh new look at risk and has endorsements by some of the top thought-leaders in project risk.
Take a look at the page here http://www.projectacumen.com/products/acumen-risk/ and let me know if you have any questions.
Tom Polen
If you look for add-in software remember that it will not produce reliable results for projects with limited resources.
If you have unlimited resources any is good, look for speed and convenience.
Email: PPadmin@planningplanet.com
Replies