For These Reasons, Don’t Calculate Availability from Work Orders

During my work experience, the source of data for calculating the asset availability was discussed several times on different levels and in many meetings.

In most cases, I found that there’s a common belief that the maintenance work orders in CMMS are the data source for calculating the availability, considering that the duration between actual job start (date and time) and actual finish (date and time) represents the asset downtime.

If we refer to the below figure (published by SMRP), it shows clearly that the above-mentioned time (actual start to actual finish) recorded on the work order is the asset “repair time” NOT the asset “downtime”, and as you can see, asset repair time represents a small portion of the asset downtime and hence can’t be considered for calculating the asset availability.

As you know, the simplest formula for calculating the availability is:

Availability % = Uptime ÷ (Uptime + Downtime) x 100

Where:

Uptime = Total Available Time – Downtime

Downtime = Scheduled Downtime + Unscheduled Downtime

Scheduled Downtime is the time required for completing the planned maintenance (PM, PdM,etc.) and unscheduled downtime is the time an asset is down for repairs or modifications that are not on the maintenance schedule.

The question now is, if the maintenance work orders in the CMMS are not the correct data source for the asset downtime, where can we find it?

The answer is that the only accurate source for the actual downtime of an asset and its category (i.e. for PM, PdM, CM, BD, Standby, etc.) is the operations daily asset operating time log.

Operations normally have a daily log for each main asset that records the categorized downtime plus the uptime.

Table 1 shows an example of a simple daily asset operating time log for one month that is updated on daily basis by Operations:

From this table, the availability can be easily and accurately calculated for each asset, as all required figures for the availability calculation formula are available and accurate.

The repair time on the work order not only can’t represent the asset total downtime but also in many cases is not accurate even for the repair time itself, as most likely (but of course this is not a good practice) the technician/trade will fill this time in the work order after going back to the workshop and by that time they will not be able to remember exactly what time the job started and finished.

Asset availability also can’t be calculated automatically from the SCADA system because the SCADA system will not categorize the downtime according to its reason; it’ll show only when the asset was up and when it was down/tripped but will not show the type of maintenance that was carried out on the asset while it was down. Also, SCADA system will not show the operations preparation work to put the asset back to service, which is part of the asset downtime.

In brief, I believe that the only accurate source for calculating the asset availability is the operations asset operating time log NOT the work orders.

I hope that this article has helped you.