Phil, My answer may be a case of locking the stable door after the horse has bolted so apologies in advance if it is. On my present site we use a part of the security swipe card system which automatically produces a monthly report on manhours burned for staff and contractors, after which a simple adjustment is made to account for breaks. Even smoke breaks are accounted for as it is a de-match site so we are confident that we get an accurate representation each month. Similar systems were used in my previous role on LNG terminal contruction sites with a corresponding system in place at the head office to capture manhours burned during the design and planning stages. It may be something you would like to consider in the future. Hope this is of help.

Phil I used to do this in a previous role and the answers were: 1. For those who clocked on and off and were "hourly paid" the figures including overtime were very accurate and derived from the payroll department's hours worked figures. For salaried employees we simply took the days worked (again from HR / payroll who knew how many sick days and holidays there were) and multiplied by the contracted staff working hours. Yes this was probably an underestimate but we had no other reasonable way of doing it. 2. We did not include sub-contractors in this although we did keep a separate track of their accidents. 3. Assumptions as stated above were based around the salaried staff working contracted hours only. You could always attempt to gauge what % of time staff work above this and add a correction factor to your figure each month I suppose but we didn't.

Hi Phil, I can completely understand why you want to calcualte this as it is still pretty much the standard benchmark. Also forgive me if I am teaching a grandmother to suck eggs but watch out for different mutipliers. there are three in common use - 100,000 hours ( a working life time); 200,000 hours (100 employees for one year) and 1,000,000 (a nice round number ). For your specific questions 1. How are accurate actual hours worked captured in organisations where many persons are salaried? This often depends on your IT systems. We collect actual hours worked. But if we don't have this information we use a calculation based on number of FTEs (Full Time Equivalents, rather than straight headcount which may have a number of part timers.) In this case you have a choice use contracted hours typically 7.4 per FTE, alternatively you can use a number which takes into account the percentage of absence that you will always have through absenee etc. 6.5 is fairly typical. 2. Are accidents and hours worked by sub contractors usually included - again how are hours worked accurately determined? This depends on what you are after. We capture and report both individually and combined, but given that you are trying to drive action you should choose your data set to send the messages that are most useful. For our long term contractors we set up contracts that requires them to report acidents and hours worked each month. If you don't have these arrangements then this can be a difficult area as you typically get accidents but not enough detail sof the hours worked. It is also common to count agency staff as employees rather than contractors as you directly direct their work. 3. If assumptions about hours worked and approximations are used what are they? I hope that the previous bit answers this I hope that this helps. Best Regards Neil