Trial and human error

C&I Issue 3, 2012

In 2011, the US Chemical Safety Board finally published the results of its investigation into the explosion at Bayer CropScience, West Virginia, in August 2008, which killed two employees. Perhaps not surprisingly, it cited multiple human failures as contributory factors. After all, it is a commonly-quoted statistic that human failure is responsible for over 80% of all major accidents.

Human failure has been a hot topic in health and safety circles for several years. Any chemical or processing operation serious about safety should already have in place a process safety management system that takes into account the various different types of human factor. At Bayer CropScience, failures included non-calibration of equipment, inadequate training of frontline personnel, faulty equipment, failure to fill a residue treater with solvent and heat to minimum safe operating temperature before feeding in a methomyl solvent mixture, among others.

In the past, it has been tempting to point the finger of blame squarely at an operator or technician. However, analysis of root causes of accidents or incidents shows there is very rarely a single reason. Generally a series of slips, lapses, mistakes or violations culminated in the final outcome, often attributed to poor communication, inadequate training, poor design, incorrect replacement parts or pressures of work on employees.

Couple this with the fact that under the Corporate Manslaughter Act, companies, rather than individuals, are liable for prosecution, and it makes sense for companies to take a more holistic approach to process safety and human behaviour, starting from the top.

The fact that human factors are closely linked to process safety means they are often viewed solely from a regulatory perspective – for example, how is an individual’s response or behaviour going to impact on plant safety? Indeed, the identification and control of risks relating to human behaviour is more pertinent now than ever. With the current state of the economy, operators are increasingly pushing plant and equipment beyond its recommended lifespan in order to remain competitive in the face of challenges from the Far East and US.

The good news is that it is possible to minimise risk and satisfy legislation, whilst at the same time making a business more efficient, productive – and profitable. The answer is to combine compliance with Lean/Six Sigma principles.

At the heart of Lean is the desire to achieve more with less by the continuous elimination or reduction of waste. Six Sigma strives to eliminate defects and any variation associated with the process in hand. A defect is classed as anything resulting in customer dissatisfaction – this can apply to product quality, design, processing, speed of despatch, paperwork etc.

Whereas traditional risk assessment relies on data, the Lean approach combines data with a more personal approach, often gathering a cross-functional team to get to the root causes of why a process or action is not functioning as it should. Asking operators and staff at the coal face what they do and why things goes wrong is often more effective than questioning senior or technical management.

Understanding and predicting human behaviour in maintenance and emergency situations, as well as day-to-day activities, means that measures can be put in place to control and mitigate the problems. Managing identified risks is all about assurance. The management systems put in place must hold up in any given situation. At the most basic level, managers should be asking themselves: what could go wrong and what will be the impact? Are there sufficient controls in place to prevent a major accident? What does each control measure deliver - how effective is it? How can we demonstrate that the management systems are of sufficient integrity and robustness?

Error-proofing is central to the ‘continuous flow’ philosophy of Lean and the ‘zero defects’ mantra of Six Sigma. Everywhere we look, we find examples of error proofing. Medicine bottles have tamper-proof caps to prevent young children from accessing their contents; a microwave will not begin cooking unless the door is closed; three pin plugs can only be inserted into a socket a certain way; unleaded petrol has a green nozzle whereas diesel is black... In fact some car manufacturers have taken this a step further, and have made tank openings different shapes or sizes for petrol and diesel, so that it is physically impossible to fit the wrong nozzle.

They are all devices to ensure that we don’t make the wrong decision or do something that will harm ourselves, our property or the wider environment. In the high hazard industries, of course, a momentary lapse of concentration, a miscalculation in calibration, insufficient training, the wrong decision leading to faulty design, can have far greater consequences.

Several changes can reduce the risks of human error. One example is to use key switch templates, so that a key can only be switched in one direction. Physical key interlock systems can also be used to ensure the correct order of valve operation. And positioning gauges as far apart as possible reduces the likelihood of reading the wrong one.

Examples of useful error-proofing devices include guide pins, limit switches, mistake-proofing jigs, counters and checklists. It is not always easy to eliminate the possibility of making a mistake, so work instructions, cue cards and warning notices should also be employed, although these are less effective than removing the ability to make the error in the first place.

Another effective Lean principle, which can be applied to human factors risk assessments, is 5S/Workplace Organisation. Workplace organisation can not only facilitate efficiency, it can also help to avert serious accidents. It involves ensuring that the right tools for the job are always available and then putting procedures in place so corners are not cut.

There was poor workplace organisation in the control room at the Three Mile Island nuclear power plant at the time of the infamous US accident. Instruments that performed different functions but looked similar were in close proximity, controls were far away from instrument displays that showed the operation of the system and readings on other instruments were difficult to read due to light glare or were metres away from the relevant equipment on a far wall. There also appeared to be no consistency in operating systems for valves etc. Pulling one lever down might close one valve, but pulling another lever upwards could close another.

In the disaster at Bhopal, India, the reasons for failure were extensive, but lack of maintenance and the removal and non-replacement of critical equipment were key factors. Similarly, visual management systems were severely lacking, with malfunctioning and unreliable gauges and meters. Another factor was that staff were overloaded and found it virtually impossible to keep track of over 70 panels, indicators and consoles. Then there was the matter of warning and instruction signs being printed in English when the majority of the workforce only understood Hindi.

Visual management systems make for a productive environment and reduce the opportunities for errors. Visual controls enable anyone in the workplace to see the status of the process: for example, whether or not the equipment is running correctly; which product or batch is being made; at what stage is the production process; whether conditions are normal or abnormal; when new material needs to be re-ordered; plus any corrective actions needed.

The majority of ‘routine violations’ occur not because a member of staff is determined to sabotage an operation, but because they genuinely believe that what they are doing is either correct or a better way of doing something. A misunderstanding of why processes are in place and the consequences of not adhering to them in the correct order is often to blame. Appropriate training is a key factor, as is the development and implementation of Standard Operating Procedures (SOPs).

Standardisation is essential if a company is to improve its bottom line. Variation manifests itself in different ways: for example, a process cannot be repeated; there are time fluctuations in completing a task; an interpretation of a task that takes away from the best method; defective supplied parts/ raw materials; replenishment of components/materials; and a different order of work carried out by different operators. Variation decreases throughput, reduces product quality, can hinder problem-solving; can hide improvement opportunities and ultimately increases cost.

The more standardised and streamlined a process becomes, the less the margin for human error and the more efficient and productive the operation becomes. It must be made clear, though, that a SOP is not an instruction manual; operators must be trained in the task and understand the reasons why.

Some of the most successful SOPs are those that are written together with the operators themselves. This way there is no question they are not understood. It also forces operators to think about the best way to do the job within the available time. The responsibility for setting and maintaining the standard is clearly up to the operator. Standards should always reflect what actually happens, not what is supposed to happen. Another bonus is that in giving the operator ownership instils a sense of pride in the work and so greater efficiency is often achieved.

As demonstrated, many potential safety risks can be anticipated and either ‘trained out’ or measures put in place to remove the human element altogether, for example, alarm systems or automation. However, one of the exceptions to the rule is in shift handovers.

Accidents attributed to inadequate shift handover include the Buncefield disaster in the UK and the 2005 explosion at the BP refinery in Texas City, US. The quality of information bestowed by the outgoing operator can have a considerable effect on site safety. Information, particularly verbal, can be misunderstood, inaccurate or omitted. The outgoing operator has completed a full shift and will quite often be tired and anxious to get home. The situation is further complicated if it is not routine, for example, if the equipment has undergone maintenance.

Log books and handover reports are a useful method of cataloguing what has occurred during a shift and identifying any anomalies. They also speed up the handover process and make the entire operation more efficient. If information can be keyed in to a computerised system, which can be accessed by all, then so much the better. This anecdotal information can then be combined with hard data from equipment readouts to facilitate the design of change management programmes.

By applying Lean/Six Sigma principles to human factors assessments it is not uncommon to realise cost savings in excess of 40%. Satisfying legal requirements should just be a starting point. Companies should continually question what more can be done, not only in terms of process safety, but in the overall efficiency of the plant. That means in every area, from the factory floor, via the procurement department and right to the very top.

Become an SCI Member to receive benefits and discounts

Join SCI