Pharmaceutical Maintenance

The Pharmaceutical Committee of the Institution of Mechanical Engineers was formed in June 2000.  It grew out of the recognition by the Process Industries Board, that this industry needed mechanical engineers to play a full part in solving its particular engineering problems in a manner that supported the key industry drivers of: -

  • ·         Regulatory compliance (and statutory, in common with all other sectors)
  • ·         Reduced time to market in the R&D phase
  • To play a full part in the meeting the pressure to reduce costs.

An early debate was whether some areas of the pharmaceutical experience were applicable to other regulated industries, such as transport.  Maybe this is so, maybe not, but we agreed that the debate was distracting.  What is true of the regulatory regimes in our industry is that they have exceedingly stringent requirements for the documentation of validation.   Some Federal Drugs Agency (FDA) edicts even push the limits of the technology that the computer software industry can deliver, and these agencies have cross border powers.  In effect the FDA can make a production facility in a foreign country unviable, by declaring a major product of that site, unacceptable for sale in the USA, through merely the quality of its supporting documentation.


The 5-year Mission for the Committee was agreed to: -

The rest of this article will deal with the background and some of the messages that were presented at the first seminar organised by the Committee in October 2000.

Maintenance in an Outsourced and Regulated Environment

Maintenance was long seen as a Cinderella activity.  However, analyses in the process industry as a whole showed that whereas the value of a capital saving on a project could be measured in a one-off saving of millions, the cost year on year of ineffective asset maintenance was compounded and could reduce or negate the planned return on capital invested.  So more and more intellectual effort has gone into the maintenance processes. 

The first thing to get into mind is that we are not maintaining assets.  We are aiming to maintain a given functionality, sufficient for its purpose, not more and certainly not less.  This infers measurement, a topic to which we shall return. 

The next fact of life to recognise is that in common with many industries the maintenance work is being outsourced.  Whilst the merits of this can be a highly debatable topic in itself, the main message that we sought to convey is that it is not an excuse for the company’s own engineers to abdicate responsibility.  Quite the reverse.  Indeed they must be precise as to what they want in order to meet the desired (measured) operational performance and regulatory needs for compliance.  Allied to the progressive achievement of cost reduction for the facility owner and an acceptable margin for the outsourcer.


So how is all this to be achieved?  Like any structure it needs firm foundations, what are the principles on which we should build.

A thorough understanding of the Regulations and their interpretation is essential.  This is not simple and indeed most of the major industry players have person(s) whose entire role is to keep up with the burgeoning regulations and their interpretation.  Not that we disagree with the intent of the regulation.  They are there to ensure our safety when we are prescribed drugs in a personal fight against an illness.  Like today’s hot topic of airline safety, I do not mind intensive searches applied to all, as it makes us all safer and from the regulators point of view, one incident is regarded as their lapse.

Pharmaceutical Regulations are based on the impact of the asset functionality of the consumed drug.  The solution to this problem is the progressive assessment of what the intended functionality is; followed by a rigorous review of its design, manufacture, and installation and testing.  All these steps being thoroughly documented and open to review by many auditors, both internal and external to the parties directly concerned.

Next you must have a maintenance team.  A team, which must allow open discussion and be fully supportive of one another.  No one should be afraid of his or her suggestion attracting ridicule or aggressive confrontation.  Openness is easier said than done in what are multi-discipline and increasingly multi-organisational teams. 

Finally there must be measurement.  Measurements can initially be very crude and then progressively made more meaningful as the team’s understanding grows.  Measurement also implies information.  Information has a key role in achieving business goals, which can reward for the efforts that it is making in terms of compliance, profit and survival.

Team Building

Having taken part in “open and frank” discussion sessions in closed rooms until the early hours and the infinitely more healthy  “outward bound schools”, I know that there are many suggested ways to break down barriers and build teams.  Some may result in more of the breakdown than those involved with the maintenance process would wish to hear.  However, the work is specialised and one should find a good supporting consultant, examine their records with similar clients to yourselves and work with him or her.

Overall any team building process must include the organisation from the top.  Most of these training methodologies have a language of their own and shorthand descriptors of “missionary” of “mover/shaker” need to be understood by all.  Seen for what they are intended, as constructive and not simply as snide, back of hand, remarks.

The consultant will have a methodology that they have developed.  All course programmes are unique because your current team and where you would like to take them are unique.  The route between the starting point and the initial destination (sorry folks, once you have got near the first goal, it will have moved!) will use components from the consultant’s toolbox, perhaps in varying sequences, and it will not be quick.  Most teams will have a core of people who have been involved for years.  Their working rules could have been very prescriptive or even very vague, but usually they demanded conformance and did not encourage change.  Now suddenly the rules of the game are all changed.  We all know that with people, change is not a rapid process.  There will be surprises, some wallflowers will become roses and there will be casualties.  Be prepared to despatch them with sympathy and generosity or risk the team’s reactions being negative and causing the consultant’s fees and your team players time to be wasted.

Measurement and Methodology

The adage “if you cannot measure it, you cannot improve it” is simple and broadly true.  Whereas you may feel performance is now better, how do you get the doubters to agree?  There can be many measures.  Availability, product throughput, product in/out of spec. operator hours per product batch, maintainers costs, etc.  Deciding which are right could be good exercises for the team in the later stages of its development.  Also their interpretation related to time periods.  Their development over time will become a role of management.

Successful measures should be simple to collect, be easily understood and show a trend reasonable quickly.  However, the disappointment is that “quickly” may well be a minimum of six months.  Although measures can be collected by any time period from a minute to a month, they are frequently distorted by exceptions, so one needs time to even-out these exceptions, to ensure that each report is not to be covered in caveats, a feature that tends to strengthen doubts, on the whole process.  The high level measures, such as cost of maintenance per £1m of installed capital can only come into their own when two complete years’ figures have been graphed.

And there is that most important measure of all – cost!  Well it is always important to the bean counters and should be part of the performance criteria set for every manager, as money is not unlimited.  We have been involved through the methods outlined below in creating an outsourced contractual environment where every maintenance activity is defined in its work content to the contractor and to the customer through a Service Level Agreement (SLA).  Then when frequencies are changed the end is an infinitely variable fixed price contract. 

More focussed data needs to be based on the methodology adopted to plan maintenance.  Reliability Centred Maintenance (RCM) is a major methodology in this area.  It was developed in the late ‘60’s and ‘70’s to tackle maintenance problems in the commercial aircraft business.  Here again as with the pharmaceutical industry, the public’s safety, regulations to achieve it and the operational economics are intimately intertwined.

The basic question of its application is “what function(s) does the system/equipment perform”.  This can be dealt with holistically and for each subset of components.  Remember we said earlier that we do not maintain assets, but functions.  The understanding of the regulations makes it clear that some functions are vital for safety, (an issue to all engineers), business needs to produce enough product at the right cost and process needs which keep regulators off the company’s back.

Safety is a paramount issue to all engineers and involves a clear understanding of the risks involved.  Often and tragically the envelope of this understanding is extended by an accident, e.g. Piper Alpha and in this context Thalidomide.  The business need also needs an assessment of risk as well as a possible change of culture.  The philosophy of avoiding the stigma of a breakdown is engrained on many maintenance cultures.  It was perhaps the Americans who coined the phrase, “if it isn’t broke, don’t fix it”.  Yet the bicycle shed culture means that unimportant items stopping in an unscheduled manner destroys hard won reputations and is to be avoided at all costs.  Not so.  If the extraction fan in the gent’s loo breaks it is cheaper to buy a new one and it can still be done that day. 

As far as the regulations are concerned the whole industry lives in dread of a Regulatory Citation.  There are websites devoted to proclaiming them.  It also seems that the threshold of acceptance keeps rising as the worst are pushed out or up to the average performance and the average are spurred to emulate the best.  Still as we say, the prize is our family’s safety and it is our challenge as engineers.


It is now commonly accepted that discreet products (your car), and complex process facilities have a cradle to grave lifecycle and their information processes, in terms of data creation and what is needed, follow and vary with that lifecycle.

Too often there remain discontinuities on what is created during a project.  An owner’s project manager may use one set of contractors whilst the needs of the owner’s operations manager, when using another set of outsourced contractors, are not met.  The oil industry have gone a long way towards bridging this gap, by more recently contracting the original main contractor to design, build and then operate the facility and be paid for the delivered product to agreed production schedules.  This is not so much an abdication of maintenance responsibility - but a divorce.  However, the very work that led to this revolution was the recognition that Exploration, Development, Production and Marketing were all approximately equal in terms of their demands on the capital resource, that the oil company's core business, in which they had to excel in to survive was Exploration and Marketing.  Thus there results a sensible integration of the design, construction and operational responsibilities, so that there is every advantage in ensuring the retention and availability of the information collected throughout the facilities life.  Our research has shown that the most demanding information needs are those of the CDM regulations, needed when the time to decommission and dismantle arrives.

The pharmaceutical industry is also faced with collecting, storing securely and retrieving vast amounts of detailed information on every operational action that the regulator may regard as important.

  1. Was the line operating within the correct limits for the product at 12:32 on 19 January last? 
  2. Were the instruments that recorded this “fact”, within their valid maintenance period? 
  3. When they were last maintained, were the calibration instruments within their valid period? 
  4. Was the operative who did the work correctly trained? 
  1. Was his/her trainer trained? 
  1. Was the calibration instrument calibrator licensed?

A paper chase indeed, and yet the escape of electronic collection is fraught.  Arguably it is just technically achievable, but therefore still very expensive.  But like the “paperless” office of dreams, total avoidance of paper is difficult and leads one into the area of electronic records and electronic signatures (ER/ES) that could even tax the authentication and non-repudiation technologies of the military.


When our team put this seminar together we realised that we were only scratching the surface of this single topic.  We also have tackled "Containment" and will address “Cleaning” and “Water” (in its variety of qualities) before returning to more general engineering and information issues on “Why pharmaceutical construction projects are different”. 

Or are they?

Norman Harris is an independent consultant specialising in Information Management Strategies and Chair of the Pharmaceutical Committee of the Institution of Mechanical Engineers.