Suggested Changes to:

Process Safety Management Regulation

29 CFR 1910.119

 And

Risk Management Plan 40 CFR Part 68

Prepared for:

The U.S. Chemical Safety and Hazard Investigation Board

2175 K Street, NW  Suite 400

Washington, DC  20037

And Others Concerned About Safety in the Chemical and Refining Industries

 By: Peter Howell, PE

Mark V, Inc

23 Edgewood Drive

Hurricane, WV  25526

304. 757.3997

http://www.markfive.com

howell@citynet.net

 June 2002

The U.S. Chemical Safety and Hazard Investigation Board

The Chemical Safety Board (CSB) held a public meeting on May 30, 2002 in Paterson, New Jersey, to present the findings and preliminary recommendations of their Reactive Chemical Hazard Investigation and to solicit feed-back from industry, labor, and the public[1].  Their first preliminary conclusion was that reactive chemical incidents are a significant safety problem.  Everyone that deals with chemical process safety on a daily basis was already well aware of this.  The CSB however, was able to quantify just how bad the problem is.  Acknowledging that they were unable to obtain sufficient information on numerous incidents, and that they suspected that there were a lot of incidents that occurred that they were unable to find in their list of databases, they still identified 167 reactive chemical incidents in the past 20 years that claimed the lives of 108 people.  This average loss of five lives per year, which is expected to be even higher if all incident information was available to the public, is unacceptable.

They concluded that there are significant gaps in safety regulations designed to protect workers and the public from the hazards of reactive chemicals.  This fact is also well known to safety professionals,  Again, the CSB quantified just how bad the situation is.  For example, they found that over 50 percent[2] of reported incidents involved chemicals that are not covered by OSHA process safety regulations and 60 percent[3] were not covered by EPA accident prevention regulations.  Clearly, the existing regulations have significant shortcomings that must be corrected.

The CSB has done a good job of identifying and quantifying many of the inadequacies of current federal regulations.  The purpose of this report is to provide specific recommendations on the changes that are needed to the Process Safety Management and Risk Management Plan regulations to close recognized Gaps in these regulations.

The Center for Chemical Process Safety

In 1985, in response to the Bhopal incident, the American Institute of Chemical Engineers (AIChE) created the Center for Chemical Process Safety (CCPS).  The objective of this group was, and still is, to establish Recognized And Generally Accepted Good Engineering Practices (RAGAGEP) that should be followed by the chemical process industries to ensure that chemical plants are designed, operated, maintained, and managed in a safe manner.  In many cases, the RAGAGEP are developed by the CCPS.  In other cases, they reference consensus standards and recommended practices developed by other organizations, such as API, ASME, NFPA, and ISA.  Compliance with the RAGAGEP developed and identified by the CCPS will result in minimizing the potential for another Bhopal type incident, and incidents in general that could cause death, serious injury, and substantial loss of property. 

The AIChE CCPS membership consists of chemical engineers that are experts in the field of process safety and come from industry and academia.  Most of these individuals have extensive experience in the design, operation, maintenance, and/or management of chemical processes.  As a group, they are highly qualified to establish RAGAGEP.

The PSM Regulation

As a result of several high loss incidents in the latter half of the 1980s, OSHA promulgated the Process Safety Management (PSM) regulation in 1992.  This regulation is very similar to the basic concepts developed by the CCPS[4].  There are however, some significant differences.  The RAGAGEP either established or endorsed by the CCPS applies to All chemical processes.  OSHA chose however, to make the regulation apply to only the few chemicals shown on a list, and to certain flammables.  Such a regulation excludes many chemicals, and mixtures of chemicals, that could, if released, cause death, serious injury, and substantial property loss.

Coverage

It is common knowledge, and the CSB has shown, that over one-half[5] of the serious incidents that have occurred in the past twenty years would not meet the criteria for coverage as established by OSHA.  This must change.  OSHA has chosen to regulate based on a list of hazardous chemicals, which relies heavily on NFPA reactivity/stability ratings.  They have selected to include only chemicals with a hazard rating of 3 or 4.  Hazardous chemicals with a rating of 1 or 2 were excluded.  This means that only 14 percent² of the hazardous chemicals evaluated by NFPA are included in the list.  In addition, there are only 325 chemicals that the NFPA has classified out of the tens of thousands of commercially available chemicals.

An additional problem with using the NFPA reactivity/stability ratings is that they are developed based on hazards presented to firemen.  They do not necessarily address actual conditions found in a chemical process[6], mixtures, or inadvertent mixing. 

As stated by the CSB, All chemicals can be reactive; hazards arise from interactions between chemicals; and these interactions can result in the release of energy or toxic chemicals[7].  The listing method is, and cannot ever be sufficiently comprehensive[8].  There are just too many chemicals currently in use, and others are being added daily.

The best approach is to use the philosophy of the CCPS and say that All chemical processes are covered by the PSM regulation.  After all, who would intentionally design, operate, maintain, or manage any chemical process using bad engineering practices?  Such a philosophy would lead to numerous incidents and high manufacturing costs.  However, the use of bad engineering practices does happen.  This usually occurs because the employer is unaware of what the RAGAGEP are, or they know the RAGAGEP that apply and choose not to follow them for any of several reasons.  OSHA does require that RAGAGEP be used for covered processes (29 CFR 1910.119(d)(3)(ii & iii), but this may not be clear to all employers.  Unfortunately, OSHA does not often site employers for failure to comply with this requirement;  possibly because the requirements are not clearly defined.  OSHA needs to elaborate of this requirement and provide more specific guidance.

If for some reason, covering all chemical processes is not feasible, it is suggested that hazardous chemicals be identified by their general characteristics rather that by a list.  The characteristics that make chemicals hazardous are well established[9]^,[10].  Methods for determining the reactive hazards of materials are also well established[11].  The hazards of the chemicals must be evaluated under all conditions that are feasible in the process, i.e. temperature, pressure, and composition.  They must also be evaluated in any feasible combination with other process chemicals, water, air, and metals.

An abundant amount of information about the chemicals in the process can usually be found on the chemicals Material Safety Data Sheet (MSDS).  Additional information about the hazards of the chemicals can be obtained from the manufacturer, often in a comprehensive bulletin on storage and handling of the chemical.  Use of these two sources alone is often adequate for developing management systems to control the hazards associated with the chemicals in the process.  Failure to learn the hazards from these sources has contributed significantly to many serious incidents.  The CSB could not find sufficient reactivity hazard information in the open literature for less than 10 percent of the 167 reactive chemical incidents identified by them[12].  For these chemicals, additional research, laboratory tests, or sharing of information with other employers or organizations would be required.

There were some concerns expressed during the hearing about the need to include storage and handling facilities as covered processes under the regulation.  Non-manufacturing processes such as mixing, blending, packaging, and storage however, provide exposure to many of the same hazards found in a manufacturing process.  Numerous serious incidents have occurred in this sector which represents about 30 percent of the 167 incidents identified by the CSB[13].  Clearly, these operations must be included in the PSM regulation coverage.  Since these operations are usually much simpler than a manufacturing process, the amount of work required to comply with the regulation would be proportionately smaller.

Use of Recognized and Generally Accepted Good Engineering Practices

Some employers mistakenly think that if they comply with each of the specific requirements listed in paragraphs (c) through (o) that they are following RAGAGEP.  However, this is not necessarily the case.  The listed practices only form a framework of the RAGAGEP that should be followed, and not the details.  The details are found in consensus standards, recommended practices, and guidelines.  It might be possible to develop a PSM regulation that included the text of the applicable consensus standards, recommended practices and guidelines.  However, such a regulation would be unusably large and impossible to keep up to date.  The preferred approach is to make it clear that RAGAGEP are required to be used and to provide guidance on whereto find such practices.

The framework of the existing regulation needs to be bolstered in a few areas to better define what is expected of the employer. 

Reactivity Matrix

As noted by the CSB, employers must address the hazards of chemicals, including all potential combinations of those chemicals, under actual process conditions as well as feasible potential process conditions[14].  In order to accomplish this, the information about the chemistry of the process must include a reactivity matrix[15].  Such a matrix is used to identify the hazards associated with the pure chemicals as well as mixing of process chemicals either intentionally or inadvertently.  This matrix must also include interactions between water, air, metals, and any other potential reactant.  This information is crucial to the safe design of a chemical process and the performance of a Process Hazards Analysis (PHA).  This through analysis will significantly improve the current situation where about 60 percent[16] of reactive chemical incidents are attributable to insufficient information about the hazards of the chemicals used in the process.

Risk Assessment

A process hazards analysis is used to identify the hazards associated with a process[17].  It also identifies the safeguards used to control those hazards.  The primary purpose of a risk assessment is to determine if the safeguards provided to control a hazard are adequate[18].  It also determines the number of Independent Protection Layers (IPL) of protection required and the Safety Integrity Level (SIL) for the safeguards.  Therefore the use of a PHA alone, as currently required by the text of the PSM regulation, but not the intent, does not ensure that a process has adequate safeguards.  The PSM regulation needs to make it clear that a risk assessment is to be done as a part of the PHA.

Contractors

Paragraph (h) of the PSM regulation is intended to apply to all contractors that work on or with the process, but the majority of the requirements appear to be directed towards contractors that do construction.  Since this is not the intent of the regulation, the text needs to be changed so that it is clear that the regulation applies to those contractors that design, operate, or maintain the process, as well as toll manufacturers.  These latter types of contractors need additional information on the hazards of the process than are listed in the regulation.  The auditing requirements of both the contractor and the employer are also more rigorous than what is implied by the current text.

Management of Change

The PSM regulation recognizes that inadequately evaluated changes to the process, equipment and procedures can lead to serious incidents.  It is not clear however, that they recognize the importance of evaluating the impact of changes to process personnel on the safety of the process.  There have always been changes in personnel occurring in the chemical process industries due to expansions, retirements, transfers, layoffs, etc.  The reorganization trend started about twenty years ago and has continued to accelerate as employers reduce staff and try to due more with fewer resources.  This has led to the remaining employees having greater responsibilities.

It is critical to the safety of the process that for any personnel change, the responsibilities of the leaving employee be determined, and that those responsibilities be transferred to other personnel⁴.  In the case of the elimination of a position, some or all of the design, operation, maintenance, or management responsibilities may be eliminated.  However, the safety and health responsibilities of the leaving individual must still be identified and transferred to a remaining employee.

Simply transferring the responsibilities to another individual is usually not adequate.  It must be determined that the individual to whom the responsibilities are transferred has the time and resources needed to meet the demands of those responsibilities, and that the individual has been adequately trained.  To the extent possible, the training must be done before the individual assumes the additional responsibilities.  It is also critical that the individual is made aware, in writing, that new responsibilities are being given to him or her and specifically what those responsibilities are.

Cost of Compliance

Some employers and trade associations have complained that compliance with the PSM regulation is too expensive;  the regulation just costs too much, or puts them at a competitive disadvantage.  In effect, they are saying that they do not want to be required to follow RAGAGEP because it is too expensive.  Enlightened companies however, have found that compliance with the PSM regulation and following RAGAGEP provides an excellent return.  Anecdotal evidence shows that employers that embrace the PSM regulation, and apply it to all portions of their operations have lower manufacturing costs, higher on-stream factors, fewer incidents, and less severe incidents.  The use of RAGAGEP and compliance with the PSM regulation in these companies is considered an investment that provides a good return in dollars, employee productivity, and goodwill throughout the community.

Employers also show their support of improving safety and health by allowing their employee experts to participate in the development and updating of consensus standards, recommended practices and guidelines.  For such employers, and trade associations of these employers, to state that the PSM regulation and following RAGAGEP can not be justified has no basis in fact or actions. 

Summary and Recommendation

The PSM regulation is good for the chemical process industries, employees, the environment and the communities around these industries.  It helps to ensure that the chemicals stay inside the pipes and equipment where they belong, and that the employees return home at the end of their work day in no worse condition than when they reported to work.  It significantly reduces the risks associated with the operation of the facility, and the potential for incidents.  The costs of incidents can be enormous.  An incident that causes $10 million of property damage can easily result in $100 to $500 million in legal expenses and personal injury claims, and loss of production and the resultant loss of income.  Over the past few years, several employers were forced out of business because of the high costs associated with an incident.  Following RAGAGEP and the PSM regulation can substantially reduce the potential for such incidents.

The PSM regulation, however needs to be amended to provide adequate protection to the employees of the chemical process industries.  At a minimum, the regulation needs to cover processes using all hazardous chemicals.  The use of a list to identify these chemicals is not comprehensive and can not be made so.  The best approach is to list the characteristics of hazardous chemicals and cover any process that uses chemicals with those characteristics.

The PSM regulation requires that employers use RAGAGEP but does not provide adequate guidance on just what is required.  A few changes are required to the text to correct this situation.  Changes that are need to ensure that all hazardous chemicals are covered and changes that clarify that RAGAGEP must be used are included in this report.

The incorporation of the suggested changes to the chemicals which would cause the process to be covered would include all of the chemicals involved in the 167 incidents identified by the CSB.  In addition, compliance with a PSM regulation that incorporates the clarifications suggested would have prevented these incidents.  These two facts show that the suggested changes must be incorporated into the PSM and RMP regulations if OSHA and the EPA truly want to protect the employees of the chemical process industries, the public and the environment.

None of these changes are likely to happen unless the White House and Congress change their priorities.  The U.S. Chemical Safety and Hazard Investigation Board, professional and trade organizations, and the public can make this happen if they let their concerns be known.  Today is a good day to write your letter.