Assistant to the Chief Engineer The Boiler Inspection and Insurance Company of Canada
Summary: The following article is a part of National Board Classic Series and it was published in the National Board BULLETIN. (9 printed pages)
Pressure vessels with quick-opening doors command a great deal of respect from the Boiler Inspection and Insurance Company of Canada (B.I.& I.). We know by simple calculation that a pressure vessel eight feet in diameter with working pressure 150 psi, has an end force of just over one million pounds acting on the heads. This does not worry us unduly in a conventional pressure vessel with welded heads. However, when the head serves as a door and it needs to be held in position by some form of locking mechanism, we have to be more concerned. Brick curing autoclaves, metal bonding autoclaves, autoclaves used in the textile industry, wood treating cylinders, tire vulcanizers, even sterilizers can all be likened to a loaded rifle with the hammer cocked when in operation. It doesn't take much to set it off. It's a lethal weapon and it needs to be handled very carefully.
Our respect for pressure vessels with quick-opening doors does not stem from theoretical calculations alone but from bitter experience.
- Some 12 years ago, an autoclave failure occurred at an insured concrete plant in Hamilton, Ontario. An 8'6" diameter by 108' long brick hardening cylinder, one of six installed in the plant, had the door blown off just as the steam pressure reached its normal maximum 145 psig at the start of a curing cycle. After deflecting off the low wall of the loading bridge pit, the door of the autoclave pierced another wall and continued piercing the plant research laboratory located over the steam kilns. The walls and a large portion of the plant's roof collapsed from the pressure of the explosion.
The 45 ton autoclave moved 150 feet away from its foundation and destroyed a delivery truck, curing racks, and numerous cubes of inventory block. Fortunately and miraculously no one was seriously injured.
The incident was fully investigated and the cause of the door failure was attributed to reduced bearing area of worn wedges and possible slippage of the locking ring.
In June 1969, another concrete block manufacturing plane in Hamilton was leveled by an autoclave door failure. The B.I. & I. was not involved in this accident but we had previously insured the location and were naturally shaken by the disaster.
The newspaper report on the incident gives the size of the autoclave as 12 feet in diameter by 80 feet long and indications are that the door blew off during operation. The roof of the plane was lifted 10 feet into the air, one wall was knocked down, and steel plates and concrete blocks were hurled 150 feet into the air. The vessel tore free of its concrete supports and ripped across an empty lot into a children's club (thankfully the club was empty because the children were away swimming). The club was flattened and the autoclave continued for another 200 feet squashing three pick-up trucks and two cars before slamming into a wall of an autobody shop. The wall fell on top of at least four parked cars and the autoclave finally butted a Volkswagen 80 feet into the middle of a nearby street.
Unfortunately, one operator died instantly and three others were reported to be critically injured.
We do not have any details as to the cause of the accident but photographs 3 and 4 attest to the devastation. Note in both cases the plants appear to have been bombed. They probably were, if you accept the fact that the energy stored in autoclaves of this size is approximately equivalent to about 80 pounds of TNT.
Following these disasters, a special Ad Hoc Committee on Autoclave Safety was set up by the National Concrete Producer's Association. C. A. Williams, B. I. & I.'s vice president of engineering, represented the insurance industry on this committee that was responsible for the 1970 publication of the pamphlet entitled "Safety Precautions for Autoclaves." The foreword of this booklet is of special interest.
"In the past 20 years (1950-1970) about 200 concrete block manufacturers in the United States and Canada have installed autoclaves to high-pressure steam cure their products. Prior to 1950 there were about 20 autoclave plants in operation. In these 220 plants employing this curing process there are probably well over 500 autoclaves in use today.
In the history of the concrete block industry, there have been eight autoclave failures; one each in New Mexico, Texas, North Carolina, Virginia, Illinois, Quebec and two in Ontario. Seven of the eight were quick-opening door failures, and six of these seven were wedge-lock type doors.
A recent tour of autoclave plants in Canada pointed out major deficiencies that effected the safe operation in three of the twenty-one plants inspected. Most common deficiencies found were related to the autoclave door safety locking device.
Due to wear and tear and to a degree of over-confidence often associated with long term trouble-free operation, most of the autoclave doors should be thoroughly inspected NOW! Subsequent inspections by a representative of the door manufacturer or other qualified person should take place every three years. Weekly, monthly, and yearly inspections are the responsibility of plant personnel and insurance company inspectors."
A copy of this pamphlet was issued to each inspector in the B. I. & I. with instruction to read and study the material and retain for reference. The need for adequate overlap on wedges was stressed to all inspectors.
We do not know of any major disasters in Canada since 1970 but we have, from time to time, heard of incidents involving autoclaves that serve as grim reminders of the hazards involved in operating these vessels.
On December 3, 1974, an operator was killed at an aircraft plant in Brampton, Ontario, when the door of a metal bonding autoclave (an 8' by 30' vessel) was blown open by residual pressure as the operator was about to open it. The operator was slammed against a concrete wall and died instantly.
Newspaper reports on the inquest into the death of the operator indicated that the vessel was being operated without a safety device that had been recommended by the insurance carrier.
In April 1976, a 17-year-old youth was killed in a cindercrete block manufacturing plant in Regina. This incident involved a 10-1/2' diameter by 124' long autoclave with a Ring-Lok Quick Opening Door. The autoclave was being vented at the end of a cycle and pressure was down to 5 psig on the gauge when the young man, impatient at the delay in waiting for the pressure to dissipate, decided to open the door despite shouted warnings from a senior operator. The door flew open, striking the concrete wall and killing the youth.
On February 2, 1977, the door of a tire curing autoclave 12' 8" diameter by 9' long blew open at a plant in Montmagny, Quebec, shortly after it was placed in service. It completely demolished a 50' by 200' building in which it was located. Fortunately no one was injured.
In May 1977, a vessel used for curing retreaded tires blew open at a location in Toronto. This vessel bears little resemblance to the vessels mentioned previously except for the fact that the cover is held in place by a concealed locking device and is therefore analogous to the loaded rifle. No one was hurt and the vessel, which was badly damaged, was replaced by the manufacturer at their own cost. At about the same time a similar failure occurred in Alberta, and again no one was injured but building damage was substantial.
In August 1977, at a plant in Edmonton, the chief engineer and production superintendent entered an 8'6" diameter by 140' long brick hardening autoclave to examine the rails which carry the loading cars. While they were inside, the operator, unaware of their presence, proceeded to push the tram of loaded cars into the autoclave. The two men managed to climb on top of the stacked blocks and began crawling desperately along the 14 inch space above the blocks toward the front, but were carried inside by the train so they were 80 feet from the door when the train stopped. They got to within 20 feet before the operator closed the door, and their shouts could not be heard above the din of the forklift truck that was used to push the train into the vessel. As steam was being admitted, the men managed to remove some blocks and pull free a steel tray that they used to pound on the door.
A welder working nearby heard faint thumping sounds and mentioned it to the operator. But after both men listened near the door they could hear nothing unusual. However, they decided to shut down the unit and open the door anyway. Both men were unconscious but quickly revived. They had only suffered numerous cuts and bruises in their desperate scramble over the blocks.
Why do these vessels fail? In almost all cases it is a combination of improper operation and poor maintenance.
How can we ensure safe operation of these vessels? The answer is fairly obvious.
- The vessel must be designed, fabricated, and inspected in accordance with Section VIII of the ASME code.
- The vessel must be properly installed on adequate foundations with provisions for thermal expansion.
- The vessel must be properly operated. This is the responsibility of the owners or managers. Their attitude is all important. They are the ones who should ensure that operators are fully trained and re-trained for the job.
- The vessel must be properly maintained;
- The vessel should have the required complement of safety devices.
The human element cannot be overlooked in the prevention of accidents. All too often after a long association with an operation, even key operating personnel have a tendency to become careless or lax. Personnel without complete knowledge of the hazards involved in the operation are even more likely to become careless and fail to follow safe operating procedures. Thorough training of operating personnel is most important and special attention must be given to the following:
- A complete explanation of the entire process and hazards involved.
- A thorough understanding of routine duties and emergency procedures.
- Thorough instruction and explanation of the functions and operation of control devices.
- An explanation of how safety devices function.
- Proper procedures for depressurizing the autoclave before opening the door.
- Proper procedures for operating the quick-opening door.
- The dangers involved by forcing operating mechanisms.
- Instructions to report at once any malfunctioning of equipment or control devices.
- Instructions should include where to look for wear and evidence of possible failure of equipment.
It is fairly certain that the autoclave disaster in Hamilton stemmed from the fitting of an oversized gasket which prevented proper overlap of the wedges and engagement of the mechanical interlock. To overcome this problem the operator took a hammer to the interlock, bent it, and got it into position with disastrous results.
Though seemingly unrelated to autoclaves, there is a film that deals in depth with the causes of an air disaster in Paris, some years ago, involving a DC-10 aircraft. The crash was caused by a cargo door blowing open shortly after takeoff. The point in common with autoclaves is that the door closing mechanism carried a small flap which sealed a hole in the fuselage when the door was locked. The aircraft could not be pressurized unless the hole was sealed and the flap served the same function as an interlock on an autoclave. On that fateful day in Paris, the door locking mechanism did not engage properly but the operator managed to get the flap into position by applying a little extra force which bent the mechanism sufficiently and sealed the hole. Some 350 lives were lost.
A lack of adequate knowledge and understanding is dangerous, especially when working with vessels under high pressure. A very important lesson to be learned from past explosions is that providing properly constructed equipment alone will not prevent accidents. Instructions must be understood and employees must be adequately trained if potentially serious accidents are to be prevented.
Management should also establish definite operating procedures and have them prominently displayed and followed by all plant personnel. These procedures would ensure:
- that special care be exercised in loading to prevent damage to the vessel, door gasket, and gasket bearing surfaces;
- closing of the door must be done only by authorized personnel thoroughly acquainted with the locking mechanism and safety devices;
- that the gasket and contact surfaces must be clean and free of any debris. Should any binding occur during closing, the trouble must be determined and corrected. Under no condition must the door or locking mechanism be forced into position;
- that safety devices, controls, and interlocks must be checked for proper operation;
- at the end of the opening cycle, no attempt must be made to open the door until the operator is CERTAIN that all pressure has been dissipated.
Since most of the accidents on vessels with quick-opening closures are caused by improper operation and poor maintenance, we must make certain that neither is neglected.
Safety devices are dealt within the ASME code and we do not dispute the merits of the code.
However, it should be noted that the code itself recognizes that it is impractical to write detailed requirements to cover the multiplicity of devices used for quick access, or to prevent negligent operation or the circumventing of safety devices. We, as an insurance company, have a definite preference for the inclusion among safety devices of a positive mechanical locking device that when in place ensures that the door is properly positioned and cannot be opened and when released provides visual and audible warning of any residual pressure in the vessel.
Thus far, we have dealt with the subject of ensuring safe operation of vessels with quick opening closures in terms of action required by owners and operators. What part does the B. I. & I. play in this? What do we do when we are requested to provide insurance coverage on such vessels?
Our inspection procedures call for a critical inspection of the vessel when coverage is first provided. The company sees the object through the eyes of our field inspector who must verify that the vessel is constructed to code, is free of defects, is properly installed, etc. Special attention is given to the door if it is of the quick-opening type.
The inspector describes and provides a sketch of the door and reports on the condition of wedges, wedge overlap, (if a wedge type door), condition of hinges, alignment, etc. He describes or provides a sketch showing all mechanical safety locking devices that serve to prevent accidental opening of the door under pressure and any audible or visual alarms which are fitted. He determines if the operators are competent, and if they are aware of the hazards involved. He looks at the attitude of the management and of the plant. Are they aware of the hazards? Have they issued safety instructions? Have they properly delegated responsibility for the safe operation of the vessel?
In most cases, approval of coverage hinges on the presence of a mechanical safety locking device. Our philosophy is that if an operator can see, feel or hear steam or air blowing from a vent pipe immediately after disengaging a mechanical interlock, his own common sense should tell him or her to wait until the pressure has dissipated. The device buys time. It gives them a chance to think twice.
Pressure activated devices and electrical devices do fail despite the many guarantees issued by the designers. Operators all too often regard such devices as being infallible and look to them or protection if they, the operators, make a mistake.
We feel that safety must not be compromised and cannot be overdone.
The engineering department of the B. I. & I. will not approve insurance coverage of autoclaves and similar vessels unless they are fitted with an approved mechanical locking device.
A well-known manufacturer had this to say about safety devices:
"There is a need for quick-opening doors with safety devices which remain effective even when personnel operating the door will not concern themselves about safety. We believe that, to be effective, a safety device must:
- be easily operated so that the operator has no reason to object to its use;
- be difficult to circumvent or make ineffective;
- be impossible to circumvent in such a manner that it can quickly and easily be returned to its original condition, say, at the end of a shift;
- be mounted so that supervisory personnel walking by the door can see at a glance that the device is ensuring safe operation;
- be simple so that the principle of operation is obvious and rugged so that confidence in its effectiveness is maintained."
The manufacturer goes on to say:
"The patented safety devices on our doors fulfill these conditions. The relative locations of the parts are fixed in our shop before shipment to ensure a safe overlap of wedges and no means of adjustment is provided. All parts are welded permanently in place. The bolt operates between lugs on the door and breech ring so that any maladjustment of the hinge would not affect the overlap ensured. The proportions of the safety device parts are adequate to resist even the full force of the hydraulic blinders tending to unlock the breech ring. As offered, the safety device should provide complete protection in that a two inch gate valve can only be closed when the door is safely locked, and the door can only be unlocked when the valve has been fully opened, and a hand wheel located in the path of the discharge from the valve has been operated.
"Various other means of providing safe operation can be envisaged but we believe that any involving electrical switches and circuits can be circumvented without difficulty and, therefore, cannot be considered prime protection. However, if required, we can supply at extra cost, in addition to our standard safety device, any equipment which may be considered to offer useful secondary protection. For example, a limit switch may be mounted on our safety device and arranged to operate only when the bolt is fully engaged; this switch could be used to control admission of steam or hot oil to the autoclave or to control the blow down valve. Also, a pressure actuated switch could be mounted on the autoclave and used to control operation of the hydraulic power unit, but the value of such an arrangement is limited by the degree of sensitivity of the pressure switches."
Finally, a few quotes from an article copied from the Autumn 1974 issue of Vigilance - The Journal of National Vulcan Engineering Insurance Group Ltd. of the United Kingdom. It is quite amazing how their experience closely parallels ours.
- "Pressure vessels with quick-opening doors such as autoclaves, sterilizers, and vulcanizers have long been subject to the danger of explosion and from time to time doors have been blown off for various reasons such as incorrect closure, worn parts of locking devices, and improper operation."
- "When the door is open at the end of the cycle, there may be residual pressure in the vessel which is not detectable by the pressure gauge. For example, a pressure of one pound f per in2 produces a thrust of two tons on a door two meters in diameter. In these circumstances, the person opening the door may be struck and seriously injured or killed."
Let us digress for a moment and refer to the instruction manual for the Bandag type autoclaves used for curing retreaded tires. The operator is advised as follows:
Open the manually controlled chamber exhaust valve (i.e. the mechanical interlock) after the curing cycle has been completed. Never try to open the door until the chamber pressure gauge reads "0." As a double check, place your hands around the silencer of the exhaust valve to make sure all pressure had been exhausted. Even one or two psi of air over the entire surface area of the door will create enough thrust to cause damage to equipment or personnel.
The safe operation of autoclaves depends upon the following measures. Failure to implement any of them may give rise to an explosion.
- The provision of appropriate interlocking safety devices for the door.
- Periodical thorough examination of the vessel and its fittings.
- Arrangements for regular systematic inspection and maintenance.
- Conformity with proper operating procedures.
- Adequate training and supervision of operators.
INTERLOCKING DEVICES FOR DOORS
A test cock or other equivalent device should give an audible and visual indication of internal pressure in the vessel. This cock should also be interlocked with the door locking mechanism, in such a way that the testcock has to be completely open before the door can start to unlock.
B. L. Whitley, Director
Boiler and Pressure Vessel Division
North Carolina Department of Labor
Apparently, many owners/operators and inspectors are not aware that after many months or possibly after a few years of cyclic operation, the locking mechanism on quick actuated closures and other similar devices can fail with catastrophic results without proper maintenance and care.
A close inspection of the items making up these closures include pins, bushings, bearings, bolts, nuts and wear rings. In many circumstances, these items reveal excessive wear that under certain conditions can cause the vessel to be blown open while it is under pressure. This results in injury to personnel and an untold thousands of dollars in property damage.
This problem is not unique nor is it confined to any one specific design. It is common in all types of closures having moving parts that are subject to wear and that have not been properly maintained.
Each inspector should be cautioned to examine every quick actuating device for excess wear. Surfaces of small critical parts, such as pins, bushings, bearings, etc., should be scrutinized closely for any existing abnormal conditions such as looseness, excess wear, or improper fit.
Should any of these conditions be observed, they should immediately be brought to the attention of management and the proper authorities. The inspector should also be reminded that the pressure sensing device designed to prevent the vessel from being inadvertently opened should under no circumstances be bypassed. If an inspector discovers the pressure sensing device has been bypassed, the inspector should immediately report the condition to the proper authorities.
Editor's note: Some ASME Boiler and Pressure Vessel Code requirements may have changed because of advances in material technology and/or actual experience. The reader is cautioned to refer to the latest edition of the ASME Boiler and Pressure Vessel Code for current requirements.