scope:

Scope and object

This technical report applies to analyser systems.

Purpose of this technical report

This technical report is written with the intention of providing an understanding of analyser maintenance to individuals from a non-engineering background. It is also designed as a reference source to individuals more closely involved with maintenance of analytical instrumentation, and provides guidance on performance target-setting, strategies to improve reliability, methods to measure effective performance, and the organizations, resources and systems that need to be in place to allow this to occur.

Effective management of on-line analysers is only possible when key criteria have been identified, and tools for measuring these criteria established.

On-line analysers are used in industry for one of the following reasons.

Safety and environment

One category of analysers are those used to control and monitor safety and environmental systems. The key measured parameter for this category of analyser is on-line time. This is essentially simpler to measure than an analyser's contribution to profits but, as with process analysers applied for profit maximization, the contribution will be dependent upon the ability to perform its functional requirements upon demand.

Asset protection and profit maximization

On-line analysers falling into this category are normally those impacting directly on process control. They may impact directly on protection of assets (for example, corrosion, catalyst contamination) or product quality, or may be used to optimize the operation of the process (for example, energy efficiency).

For this category of analysers, the key measured parameter is either the cost of damage to plant or the direct effect on overall profit of the process unit. Justification as to whether an analyser should be installed on the process may be sought by quantifying the payback time of the analyser, the pass/fail target typically being 18 months, although it should be noted that the contribution of the analyser to reduction in the extent of damage to, or the profit of, the process unit is difficult to measure. However, this contribution will be dependent upon the analyser's ability to perform its functional requirements upon demand.

This technical report focuses on the cost/benefits associated with traditional analyser maintenance organizations. In a modern set-up, the complexity of analysers demands on occasion data from chemotricians and scientists who may be owned by other parts of the organization, and, as such, care must be exercised to include their costs.

Questions that need to be addressed

When considering on-line analyser systems and their maintenance, the following list of key points is useful in helping decide where gaps exist in the maintenance strategy. Additionally, a structured mechanism by which the "health" of an analyser organization can be appraised is provided in Appendix 5.

1. What is the UPTIME of each critical analyser? (Do you measure UPTIME and maintain records? Do you know the value provided by each analyser and therefore which ones are critical? Do you meet regularly with operations ("the customer") to review priorities?)

2. What is the VALUE delivered by each analyser in terms of process performance improvement (i.e. improved yield figures, improved quality, improved manufacturing cycle time and/or process cycle time, process safety (for example, interlocks), environmental importance)? (Is this information readily available and agreed to in meetings with operations? Is the value updated periodically?)

3. What is the "utilization" of each critical analyser - that is, if the analyser is used in a control loop, what percentage of the time is the loop on manual due to questions about the analyser data? (Do you keep records on the amount of time that analyser loops are in automatic? Do you meet regularly with operations to review the operators feelings about the "believability" of the analyser data?)

4. Do you have a regular preventive maintenance programme set up for each analyser which includes regular calibrations? (Does the calibration/validation procedure include statistical process control concepts - upper/lower limits and measurement of analyser variability (or noise)? Is the procedure well documented? Do you conduct it regularly? Even when things are running well?)

5. Do you have trained personnel (capable of performing all required procedures and troubleshooting the majority of analyser problems) who are assigned responsibility for the analysers? (Do the trained personnel understand the process? Do they understand any laboratory measurements which relate to the analyser results?)

6. Do the trained maintenance personnel have access to higher level technical support as necessary for difficult analyser and/or process problems? (Do they have ready access to the individual who developed the application? Do they have ready access to the vendor? Can higher level support personnel connect remotely to the analyser to observe and troubleshoot?)

7 Do you have a maintenance record keeping systems which documents all activity involving the analysers, including all calibration/validation records, all repairs and/or adjustments? (Do you use the record-keeping system to identify repetitive failure modes and to determine the root cause of failures? Do you track the average time-to-repair analyser problems? Do you track average time-between-failures for each analyser?)

8. Do you periodically review the analysers with higher level technical resources to identify opportunities to significantly improve performance by upgrading the analyser system with improved technology or a simpler/more reliable approach?

9. Do you meet regularly with operations to review analyser performance, update priorities, and understand production goals?

10 Do you have management who understand the value of the analysers and are committed to, and supportive of, reliable analysers?

11. Do you know how much the maintenance programme costs each year and is there solid justification for it?

Consideration of the above questions will help to identify opportunities for continuously improving the reliability of installed process analysers. Once the opportunities are identified, the following sections are intended to give guidance in achieving the solutions with the aim of

• maximising performance and benefit of installed analysers;

• achieving full operator confidence in the use of on-line analysers;

• analyser output data becoming reliable enough to be used by operators, control systems, and other users to improve plant operation versus world-class manufacturing metrics and become best-of-the-best.