PAT System Reliability

July 8, 2008
Site-Wide Analyzer System Reliability Requirements

By Gary D. Nichols, PE

In the previous article in this series, (Process Analyzer Reliability, Maintenance & Service), we devoted most of the time to preparing to incorporate a new analyzer system into an existing site and into an existing computerized maintenance database (CMMS) correctly (see Patton below); we discussed little about the site or its reliability/maintenance organization.

For this discussion, we shall assume that the manufacturing site includes several operating/production units and a mixture of chromatographic, photometric, environmental and special process analyzer systems housed in an assortment of analyzer shelters and three-sided enclosures, and field-mounted. We shall further assume that an analyzer system maintenance and reliability organization exists, comprised of a supervisor, several analyzer system technicians and perhaps a maintenance planner.  Finally we shall assume that spare analyzer parts are ordered automatically, as they are used by the analyzer technician in the field and attributed as a cost against the respective analyzer.

As early as 1971 (See Upfold below), and reprinted in 1994 (See Dailey below), Table I shows what was believed to be typical analyzer technician service time requirements for various types of process analyzer systems. The reader may divide the annual time requirement in Table I by 52 to obtain the approximate weekly time requirement per analyzer type. Table I is reproduced here for convenience, but the reader is advised to generate his own maintenance data from actual maintenance and service times when embarking upon an effort to assign analyzer technician responsibilities to individual analyzer systems and analyzer system groups.

TABLE I: Man-Hours per Year for Analyzer System Types in 1971, Republished in 1994

Analyzer System Type 

Mean (MH/year)

Range (MH/year) 

Standard Deviation (MH/year)

Boiling Point 




Gas Chromatograph 




















Refractive Index 








If data such as that in Table I is customized for the site, we can obtain an approximate picture of how many analyzer system technicians are required to service and maintain all of the analyzer systems on site reliably.

We say “approximate” because some analyzer systems of the types noted in Table I may require more or less attention than others due to the sample in which they are placed, e.g., clean gas, clean liquid, two-phase, tarry liquid, liquid or gas with suspended solids, etc.

Analyzer system service time is not only affected by the complexity of the analyzer system and the difficulty of the sample, but also by the levels of training, experience, breadth of responsibilities, management support, union rules, system for obtaining spare parts and prevailing conditions of the analyzer systems. Training, experience and breadth of responsibilities apply to supervision as well as to analyzer technicians.

Analyzer technicians are typically assigned geographically, that is, an analyzer technician is assigned to maintain all the analyzer systems within the battery limits of one or more operating units, based on the number of analyzer systems and complexity within the unit(s).  Analyzer technician assignment by geography has the distinct advantages of permitting camaraderie and cooperation between the analyzer technician and operating personnel, greater analyzer technician familiarity with unit MoCs, task-specific safety procedures, unit safety procedures, control system and associated alarms, emergency procedures, unit operating chemistry and unit personnel hazards.

Another way to assign responsibilities to analyzer technicians is by analyzer system type (GC, photometric, environmental, oxygen, etc.), but this has the disadvantages of requiring too-wide geographic coverage, too highly specialized training and causing difficulties in providing vacation and sick leave backup support.

At this point, we have assumed, we hope correctly, that a computerized maintenance management system (CMMS) exists. We also assume that analyzer technicians are assigned to specific analyzer systems in specific operating units, that the spare parts system is automated to attribute spare parts usage to particular analyzers (possibly part of the CMMS), and that spare parts are automatically ordered as they are used in the field.

The Typical Service Call

Then let us move on to what might happen during a typical analyzer-system service call (reactive maintenance).  First, the board operator notes that the analyzer measurement has been steady and unswerving for several minutes and not varying as he knows the process to vary over time. The board operator calls the analyzer technician to report the problem. The analyzer technician reports to the control room, examines the concentration and other readouts from the analyzer and agrees that the analyzer system is not responding well to the process.

The analyzer technician then officially checks in to the unit through the control room, informs herself of any safety procedures in effect at the time, completes any lock/tag/try procedures called for, dons the personal protective equipment required for the analyzer system and proceeds to the unit in question to diagnose and solve the problem.

First, the analyzer technician records the “as found” condition on the analyzer system using his laptop PC (safe for use in the area electrical classification of the analyzer system). He expects that sample conditioning system pluggage is responsible for the unvarying measurements, since he knows, based on the process chemistry, that a heterogeneous (particulate) catalyst is used in the manufacturing process, that a small amount of tar is formed and that the analyzer is on the “bottoms” of one of the product purification stills.

Therefore, the analyzer technician systematically inspects the sample conditioning system from the inlet forward to the analyzer. After some time, she discovers a mass of catalyst particles and tar in one of the needle valves that controls the balance between fast-loop and measure-loop flow, and she replaces the needle valve and the tubing leading to and away from the valve. After waiting several minutes to determine whether the sample flow returns to normal and remains balanced, he records the “as found,” “work done,” “as left” and spare parts usage on the PC; verifies that the analyzer now reports correctly in the control room; reports this to the board operator; removes her personal safety tags and signs out of the control room and operating unit.

After Repairs

After the analyzer technician returns to the analyzer shop and reports his return to the analyzer supervisor, he downloads the laptop PC to the CMMS. The nature of the problem, work done, spare parts usage, “as found,” “as left,” “out of service” duration of the analyzer and total service time (billable maintenance time using time stamps), and other designated critical information is downloaded. The billable technician time and spare parts are automatically charged against the operating unit, the spare parts consumed on the call are automatically reordered, technician time is recorded for payroll and the analyzer technician can proceed to the next job.

Similar metrics would be applied to predictive maintenance, though the time would be pre-scheduled. This procedure is idealized, but should provide some idea of the current trends in maintenance, including analyzer system maintenance. The idealized system described here is transparent as to whether the parts are purchased and warehoused ahead of time or on consignment, or whether the analyzer technicians are direct hires or contract workers. Service contracts are usually billed differently, but should be tracked so as to attribute specific labor and parts charges to specific analyzer systems so that the lifetime costs can be attributed to a specific analyzer system.

Spare Parts

Analyzer systems that have been deemed “critical” may have an exclusive and segregated batch of spare parts that are not to be used on other analyzer systems, not even when needed or when they “fit.”  Other spare parts may be available for any analyzer system or purchased from the OEM rather than the analyzer manufacturer. Any parts or complete analyzer systems that can be repaired by the manufacturer or third parties should be returned to service or designated as a “shelf spare” at the expense and discretion of the operating unit/owner.

Labor and parts cost tracking system should be used, whether the service call is from operations (reactive maintenance) or is discovered during preventive or predictive maintenance by the analyzer technician looking at the control board on in the field inspecting the analyzer system.  Preventive or predictive maintenance should have a checklist of items to be examined and tested, along with a start-to-finish safety checklist.

Periodically, perhaps quarterly, operations and maintenance should conduct a review of analyzer system performance during the previous time period. The review should focus on analyzer system performance, not on the technician or supervisor, and on what is actually delivered by the analyzer system in the unit versus what is expected of the system, and what is necessary to reconcile any differences.

We have assumed electronic record keeping, but we must acknowledge that some systems still rely upon paper information transfer. In this case, information transfer and action will be much slower and perhaps susceptible to errors, especially after the field work in completed.


  1. Patton, Joseph, “Computerized plant maintenance: Predictive maintenance prevents failures from happening at a bad time and fixes them before they cause damage,” pp. 46-53, Intech,  July 2007.
  2. Upfold, A. T., “Man-Hour Ratings for Standardized Instrument Maintenance,” Instrumentation Technology, ISA, Research Triangle, NC, USA, pp.46-48, 1971.
  3. Dailey, William V., “The Whys, Whats, Whos, and Hows of Process Analyzer Maintenance,” ISA/1994, Paper Number 94-003, ISA, Research Triangle, NC, USA, 1994.