of the PRO-M product (the basic data structure and in-service stressors).  The third involves modifications that we can easily introduce at the request of users.

Basic Structure Of The Recommendations

PRO-M makes recommendations about PM tasks and task intervals in 8 categories.  The 8 categories cover, 1) the Functional Importance of the equipment (Critical or Minor), 2) Duty Cycle (High or Low), and 3) Service Conditions (Severe or Mild), each of which provides a binary choice (giving 2x2x2 = 8 categories).  The choices are shown in the following table for a specific component type:

Functional Importance:	Critical				Minor
Duty Cycle:	High	Low	High	Low	High	Low	High	Low
Service Conditions:	Severe		Mild		Severe		Mild
PM Task Name	CHS	CLS	CHM	CLM	MHS	MLS	MHM	MLM
Visual Inspection	1Y	1Y	2Y	2Y	4Y	NR	NR	NR
Oil Analysis	3M	6M	3M	6M	1Y	2Y	2Y	NR

In this example, 1Y means one year, 3M means 3 months, and NR means Not Recommended.  Each category is designated by a 3-letter code, thus, the CHS column contains interval recommendations for this component type operated with critical functional importance, a high duty cycle, and in severe service conditions.  In less demanding conditions the recommendations may be relaxed, as shown.

This means that the same electric motor, operated under different conditions would get different recommendations from PRO-M.  Of course, the user has to know what the conditions are for this particular motor.  So PRO-M provides him, always at a click of the mouse, with a brief explanation of what is meant by a high and low duty cycle, and severe and mild service conditions, specifically for this component type.  In addition, PRO-M presents the user with a simple automated checklist to determine functional importance.

In this way, PRO-M makes initial, at-a-glance PM recommendations depending on these categories.  However, the initial recommendations are purposely somewhat conservative because it is recognized that a one-size-fits-all approach, even within one of the 8 categories will not be appropriate for all situations.  The user is informed by another mouse click of the PM tasks that are the most important and least important if his objective is to aim for a comprehensive set of PM tasks.  In these circumstances he can be sure he preserves the most important ones and consider dropping the least important.  At the same time he is informed of the single most beneficial task and of the two tasks (sometimes three) that together achieve a very significant reliability gain if he is looking for a really minimal program.  This gives him an immediate sense of how to 'sharpen' the recommendations for his purpose.

If he wants to be sure how his selection affects the failure rate, maintenance costs, and the return on investment, he can simply press a button to find out.  To run the calculation in the case of severe service conditions (a category ending in 'S', like CLS) PRO-M will ask why the conditions are severe.  The user is prompted to select any or all of six groups of in-service stressors that might apply to this component.  These stressor groups are:  Heat, Humidity, Dirt, Vibration, Fluid Quality, and Biological.

A little algebra shows that there are 62 ways of choosing combinations of these 6 stressor groups, and these are all available for the severe service condition categories, making 126 different sets of conditions that can be addressed!

In-Service Stressors

Internal to the calculations and transparent to the user, the activation of the effect of service stressors on reliability is initiated by keywords that appear in the Degradation Influence field of each component table.  The contents of this field are exactly as determined by the subject matter experts at the time the core information is gathered.  The keywords for the main stressor groups (duty cycle is treated the same way as the 6 other stressor groups) are as follows:

STRESSOR TYPE
Duty	Heat	Humidity	Dirt	Vibration	Fluid	Biological
Run time	Heat	Humid	Contam-  ination	Vibrat	Salt water	Wood-  pecker
Runtime	Hot	Moist	Dirt		PH	MIC
Start	Temp	Wet	Ash		Debris	Mussel
Normal wear	Thermal	Spray	Salt		Chemistry	Fish
Load factor		Condens	Foreign  material		Fluid	Bio
Cycle			Dust		Particle	Algae
			Rust		Particulate	Vegetation
			Environ-  ment		Two phase
					Two-phase
					Hammer
					Oscil
					Flow
					Quality
					Off-BEP
					Sand
					Silt

When, and only when, the user has indicated that a particular stressor group should be applied, a keyword search is conducted on the Degradation Influence field.  For each failure mechanism where a keyword is found in this field, a stressor correction is applied.  The general procedure when stressors act is to reduce the failure free time for wearout mechanisms, and to significantly increase the strength of random mechanisms.

Tuning The Data To Generate A New Component Type

Clearly, the stressors represent the effects of a very wide range of commonly occurring industrial conditions.  Historically, this was done so that the wide range of conditions encountered in a nuclear power plant could be adequately represented.  Yes, even in those plants not everywhere is clean and air-conditioned - conditions range from very hot and humid, to severe vibration and aggressive water chemistry, to high levels of dust, oil vapor, and salt laden air! But from an early stage, personnel from coal burning plants began using these data sets for a wide range of applications because their conditions could be represented adequately by activation of the existing stressor groups.  The same thing has happened as the data sets have been taken up by personnel who service the transmission and distribution equipment in the power delivery side of the electric utility industry.  Much of their equipment, including pumps, motors, valves, compressors, and instrumentation is exposed to the elements.

For some component types, however, the differences could not be spanned by the existing stressor types and we had to make mild additions, such as adding ash to the list of stressor keywords for the Dirt stressor, and adding Ash to the Degradation Influence field where it was needed.

In some instances, differences in operating conditions lead to the introduction of an entirely new component type as an edited copy of an existing component.  Thus, fossil plant main turbines were born as close copies of nuclear plant main turbines but with some key differences in subcomponents and degradation mechanisms.

When adding new component types for an entirely different part of the industry we have sometimes modified the existing stressor keywords as described above for existing component types, e.g. adding woodpeckers for the wooden utility pole components!

The point is it is easy and inexpensive to modify the existing stressor approach and the effects of operating conditions and component design for existing or new component types when it is judged to be necessary.  Therefore, we anticipate that this will be easily and quickly accomplished as we expand the use of PRO-M into new industries.