In some plants, the various work order categories were put into a list without a lot of thought, and they were then put into use. An example of such a list is shown below:

The trouble with a table like this, is that the selection options describe different characteristics of the work to be done. For example, in the plant which used this list, consider the possibilities if there was a serious accident involving a machine shop lathe, and a decision was made to immediately bring in a contractor to replace the lathe (a capital project). Would the correct work order code be CA, CO, EM, EQ, MA or SA? Chances are, different people would select different codes, based on which they determined to be the most important from their point of view.

Many work order "types", "classes" or other categories contain tables of this nature, and they frequently break three of the basic rules of data-base management which are:

- each field must contain a list of "mutually-exclusive" options. "Mutually exclusive" means that an informed user would be able to select one and only one option that describes the characteristic of the work covered by that field. This means that there must be a separate field for each characteristic of the work.

- the list of options must be short - about eight is a good goal, but the list must never be longer than can be viewed in large font without scrolling. I have seen lists with over a hundred options - these lists do, of course, provide no value and the item that's at the top of the list was the one most frequently selected.

- a default selection should not be used. Some thought must be given to the selection before it is entered.

Typical characteristics of maintenance work orders which can be used are:

- the limitations on scheduling the work
- the reason for the work
- the urgency of the work
- the source of funding
- the manpower resources to be used
- the current state of the work

There are other possibilities (such as "root cause of the failure" which has its own unique problems) but lets look at the above list in more detail.

The limitations on scheduling the work are normally easy to define, and the options include

• "on the run", i.e. there are no limitations
• "shutdown preparation", i.e. work which must be completed before a specific scheduled outage
• "major shutdown", probably requiring a total plant outage
• "area shutdown", requiring an outage in a defined section of the plant
• "deadline", for work which has no value if done after a certain date (e.g. painting the boardroom for a meeting with important investors)
• "emergency", i.e. it must be started immediately
• "urgent", which could be defined as work that must break into an established work plan, such as a weekly work schedule
• "by arrangement with operator", e.g. to work on standby equipment

There are other possibilities, for example some operations may shut down parts of an area to produce certain products, and maintenance may be scheduled during these times.

The reason for the work is a little more difficult, because more than one reason may apply. A typical list of options might include

• "safety", or work that is required to eliminate a hazard
• "environment", or work that protects against environmental non-compliance events
• "quality", i.e. to maintain or enhance product quality
• "service", such as lubrication, calibration, winterizing or other time-based care
• "inspection", usually time-based (and should be defined to exclude "legal" - see below)
• "improvement", which can be separated further, if desired, into "process improvement" and "reliability improvement"
• "legal", or required to comply with regulations, such as some inspections.
• "repair", or correcting problems which have been identified by some type of inspection.
• "spare parts", or work required to make or overhaul items which will be returned to inventory.

There may be other reasons for work that are appropriate for specific industries.

There are circumstances where more than one reason may apply. For example, if a pump is to be repaired because it has noisy bearings, and at the same time a new type of coupling is to be installed to improve reliability, part of the job is "repair" and part is "improvement". Some guidelines are required to assist users in this decision, and two possible options are that the part of the work which will cost the most money should determine the "reason" code, or the part of the work that is most important to be recorded should determine the code selection.

The urgency of the work (which can also be called its "priority") is different from the scheduling limitations, and defines a target time frame for work completion. Typical options include:

a) for work which does not require a shutdown:

- required within 2 weeks
- required within one month
- required within three months
- required immediately
- required within one week

The last two options do duplicate two of the codes used to define scheduling limitations, and the CMMS should be designed to ensure that the either of these code selections are automatically duplicated in the "limitation on scheduling" and "urgency" fields.

b) for work which requires a shutdown

- required at the first opportunity (including unscheduled outages)
- required at the next scheduled outage
- required at one of the next 2 (or 3) scheduled outages

The first of these options should be reserved for equipment which is an advanced stage of failure (very noisy bearings, hazardous leakage, etc) and the resulting list of work should be updated frequently and carried by people who are likely to be involved if an unscheduled outage occurs. This would include people on weekend call, for example.

One issue with the use of "urgency" codes is that the perceived urgency of a job is often very subjective or even emotional. Good predictive maintenance can make the assessment of urgency more objective, but there will always be the need to balance the importance of the wide range of jobs that are the work of a maintenance organization.

The source of funding is usually straightforward, and includes the following

• "routine expense", for most maintenance work
• "major maintenance project", if the plant uses a more sophisticated project management process for maintenance work over a certain dollar value
• "capital project"
• "warranty work"
• "inventory", if it is the plant's policy to charge repairs to spare equipment to an inventory account

The manpower resources to be used may require two fields, depending on the organization. The first should describe the trade or skill required ("electrician", "carpenter", etc) and the second the area maintenance crew from which the resources will be drawn ("finishing area maintenance crew", etc).

Resources which should be included in the appropriate list include contractors, engineers, vendors' representatives, consultants and any other resources that are used to complete any work.

The current state of the work is used for management of work orders, and is also valuable information for anyone, especially operators, who want to know the status of their work requests. Typical status codes include

• in the backlog
• prioritized and to be planned
• in planning
• waiting for materials
• ready to schedule (all materials on site)
• scheduled (on a weekly or shutdown schedule)
• in progress (some time charged)
• completed (or cancelled)

The fields used for describing the current state of the work and the resources to be used are usually well-managed. It is the other codes, for scheduling limitations, reasons, urgency and funding, which are frequently combined to one extent or another, with the result that the data in the database can be so "dirty" that it has little value for analysis or control.

One over-riding consideration in the use of any of these codes is one of value. Computers make it very easy to gather and store large quantities of data, but unless it is used to create real value, it should not be collected at all. For example, if the only "reason for the work" that is ever reviewed is safety work, because a commitment has been made to provide this to the safety committee, then the only two options in the "reason for the work" field should be "safety" and "other".

In many of these fields, the options may require a carefully-considered definition, which in turn requires that its someone's responsibility to make sure there is no abuse of the options. For example, "safety" work may be defined as "work that is required to eliminate a hazard, or work that originates as a recommendation from an accident investigation". This still requires some judgment on what a "hazard" is, but eliminates routine work, such as re-painting the lines in the parking lot, from the list of safety work.

Also, some code options may not strictly meet the objective of being "mutually exclusive", such as the "inspection" and "legal" options in the suggested table of reasons for the work. In these cases, the definition should indicate any exceptions (in this case "inspections" is defined to exclude inspections that are required by laws or regulations, such as pressure-vessel and elevator inspections).

If the above guidelines are adopted to manage work orders, clean, valuable data will reside in the CMMS. At regular intervals, the value of this information should be reviewed with a goal to maintaining as simple a process as possible.

One last word - the people who enter work order codes and have to make decisions code selection MUST have some regular feedback on its value, or they will soon lose the discipline that is needed to maintain clean and consistent information. If this value can not be demonstrated to them, then it is probably not worth entering the information in the first place.