Flammable Hazards in Sewerage Systems

(Author's name withheld)


Table of Contents

1. Introduction
2. Background - Flammable Hazards in Sewerage Systems
3. Literature Review
4. Ignition Sources
5. Hazardous Zones
6. Flammable Hazards
7. Survey
8. Results of Survey
9. Discussion
10. Conclusion
11. References


1. Introduction

On the 22nd April 1992 in Guadalajara Mexico, explosions in the sewerage system killed 190 people and injured 1470 people. The explosions occurred from accumulation of petrol, hexane and possibly methane in the system. The disaster has alerted a Sewerage and Water Provider in Australia, in which the research was carried out, to review safety procedures used when accessing sewers which may contain flammable gases/vapours-air mixtures in the explosive range.
Subsequent to the Mexico disaster, incidents occurred in the Provider's area of operation where natural gas accumulated in sewers in concentrations within the explosive range. If these explosive gas/air mixtures had been ignited, damage to property would have occurred and a serious risk of injury or death to people in the vicinity. These natural gas incidents showed that Australia is not immune from disasters similar to that which occurred in Mexico.
As a consequence of these incidents, some Providers of these services have been revising safe working procedures. This research looked at how one Provider has approached this problem. The main emphasis of their procedures was the issue of a Technical Information Sheet to the sewerage maintenance areas. The major points of the Technical Information Sheet were:

      1. "Where practicable, keep potential sources of ignition at least 3 metres away from the chamber (manhole) e.g. 2-way radios, "motorolas", cellular phones, radios, pagers, torches, other electrical equipment, vehicles and other motorised equipment.
      2. "Wet the seal with water."
      3. "Switch on gas detector and place downwind and as close as possible to the access cover."
      4. "To lift the access cover use a lifting device which maximises the distance of the person from the cover."
      5. "Slowly lift cover about 50 mm."
Of the procedures, the question as to whether the procedure "Wet the seal with water" was necessary was unknown. The Technical Information Sheet was issued as a matter of urgency, consequently not all the above steps were completely researched as to their necessity. In particular, as to whether friction between an access chamber cover and its surround could be a source of ignition when the cover was being removed. Consequently it was recommended that water be added to the seal.
One may ask why gas monitoring could not be done by means of a sampling probe passed through one of the "pick" holes of an access chamber cover. Unfortunately many access chamber covers in the Provider's area of operation do not have holes suitable for direct sampling inside the chamber. The procedures that are listed in the Technical Information Sheet assume there is an explosive atmosphere in the chamber.

 

1.1 Statement of aims

The aims of this research were to:

      1. Obtain data on fire and explosions that have occurred in sewers and associated structures.
      2. Investigate whether friction between an access chamber cover and its surround could be a source of ignition when the cover is being removed.
      3. Determine compliance by sewerage maintenance people in carrying out the procedures according to the Technical Information Sheet.

  

2. Background - Flammable Hazards in Sewerage Systems

Flammable gas/vapour hazards are an important consideration in sewerage systems because of the significant risk of fire or explosion. Not all flammable hazards in sewers emanate from the decomposition of sewage. Other sources such as trade waste discharges and leaking natural gas pipes can contribute to the most significant hazards because they are unpredictable and can involve a large amount of flammable material entering the system in a short period of time. The occurrence of flammable hazards in sewers may occur from various events. Situations leading to flammable gases/vapours in sewers are given by a number of authors (Abbott 1993, Feldman 1977, Wolfs 1959, National Fire Protection Agency (NFPA) 1982, Prentiss & Laughlin 1995). These include:
Flammable hazards from illegal trade waste discharges can occur on a regular basis e.g. monthly (verbal communication with Water Board, Sydney). A number of authors mention that methane may enter a sewer from leaks in nearby natural gas utility lines (Gerardi 1982, Feldman 1977). Incidents have occurred in an Australian Water and Sewerage Provider where natural gas from leaking natural gas pipelines have entered the sewerage system and concentrations have exceeded the lower explosive limit (LEL). In one situation gas levels remained above the LEL for approximately 10 days before corrective action to repair the leak was taken (verbal comm with Water Board, Sydney). This type of incident illustrates the need for good communication and cooperation between Providers to prevent the activities of one body from impacting on the activities of another. When things go wrong, as they inevitably will over a period of time (Tweeddale 1994), systems must be in place to respond quickly to emergency situations. The primary consideration is of course to prevent incidents occurring.
Under normal conditions in a naturally ventilated sewer, high levels of flammable gases would not be generated. However, illegal trade waste discharges or leaks of natural gas can overcome the defences of natural ventilation. Sewers are designed to operate at certain gradients and flow rates to prevent accumulation of sewage. Sewage does not generate methane until it is approximately 3 days old. The National Fire Protection Agency (NFPA) 820 (1992) mentions that the potential for ignition within the liquid stream of a treatment plant is restricted to flammable materials entering the start of the treatment process and to flammable gases generated as by-products of waste water decomposition in non-flowing or septic portions of the liquid stream. Wolfs (1959) points out that lack of ventilation in a sewer over a sustained dry weather period produces conditions that may result in the build up of hazardous concentrations of vapours from combustible liquids. A small constant source of contamination over a long period of time could be more dangerous than a large discharge. Presumably in this article Wolfs is suggesting that a small leak of combustible material may go unnoticed.
Methane is a flammable gas occurring in sewers from the decomposition of sewage or from infiltration of natural gas (natural gas is essentially methane). Methane is produced by microbial action on carbohydrates and proteins found in sewage. Under anaerobic conditions i.e. when oxygen is absent, the bacteria multiply and convert the waste to energy and methane. The gas is colourless and lighter than air and tends to accumulate initially at the top of a manhole (Gerardi 1982). Natural gas contains odorants so that it can be smelled when a leak occurs. However, in the presence of sewage odours, natural gas may not be readily detectable by smell. Depending on the pH of the sewage under anaerobic conditions, hydrogen sulphide or ammonia may also be produced (Gerardi 1982).
Hydrogen sulphide is also a flammable gas, however, it does not present a risk of fire or explosion in sewers because the air/aqueous partition coefficient of this gas is too low for sufficient amounts to enter the air space to generate a flammable atmosphere (Bowker, Smith & Webster 1989).
Ammonia/air mixture is colourless and possesses a sharp urine-like odour. The concentration of nitrogen in the form of ammonia in untreated sewage is typically in the range 10-50 mg/L (Abbott 1993). At 20oC, below a pH of 7, ammonia exists as ammonium ions in solution. As the pH is increased above pH 7, the amount of ammonia as a dissolved gas begins to increase and some ammonia is released to the atmosphere. However, even if the total amount of ammonia were released, there is insufficient quantity in sewage to constitute a flammable hazard. This can be demonstrated by reference to distribution curves of ammonia/ammonium ions in solution as a function of pH and temperature (Sedlak 1991) and then by reference to tables of the partial pressure of ammonia over aqueous solutions of ammonia (Perry 1984).
Foster (1979) mentions that any sewer passing near gas mains or gasoline storage tanks are "potential killers'. He mentions that when ventilating the sewer, blowers and the engine should be located at least 3m from the manhole and preferably cross winds so that explosive gases will not reach the engine. He stresses the prohibition of smoking and that radio equipment should not be allowed in the sewer. He places a mandatory requirement on the use of rubber boots or non-sparking shoes and non-sparking tools.
The occurrence of flammable atmospheres in sewerage systems is unpredictable and unknown until tested.

 

3. Literature Review

The object of the literature review was to meet the needs of the first two aims of the research, i.e., to obtain data on fire and explosions that have occurred in sewers and associated structures, and to investigate whether friction between an access chamber cover and its surround could be a source of ignition. Non-sparking tools were also reviewed because some respondents to the survey used a hammer and chisel to loosen access chamber covers which were difficult to remove. Sparks generated from the use of a hammer and chisel could also be a source of ignition.

3.1 Fire and Explosion Incidents

The National Fire Protection Agency (NFPA)(1982) has documented a number of flammable gas/vapour incidents that have occurred in sewer and water systems. These are detailed as follows, number 3.1.1 - 3.1.13.

3.1.1 Leakage of Flammable Liquids or Gases into Sewers - Cleveland, OH

September 11, 1953.

A tremendous underground explosion in a combination storm and sanitary sewer demolished 1.9 kilometres of pavement of West 117th Street killing one person and hospitalising 58 others. The blast threw access chamber covers high into the air and broke out large chunks of concrete pavement, some 6 metres by 3 metres in area. Gas mains and water lines were ruptured. A second blast occurred an hour later hurling more access chamber covers into the air, but no one was hurt. At least 30 automobiles were heavily damaged or demolished in the blast. The catastrophe was thought to be due to the ignition of some flammable liquid or natural gas from local underground sources in the 1.8 metre sewer line that ran beneath this street. Wolfs (1959) mentions that the cause of ignition was unknown.

3.1.2 Leakage in Distribution Piping Systems - Goldsboro, NC - April 12, 1954.

An explosion and fire in the business district was followed by a series of secondary lesser explosions at about 5-minute intervals in a nearby street sewer. Gas had entered one building through an abandoned gas pipe protruding into the basement, and fire broke out within two or three minutes and burned for about two hours before it was extinguished. A US Bureau of Mines analysis showed that the gas at the explosion site was of the same composition as that distributed by the utility and consisted of a propane air mixture. It was also revealed that the community during the previous months had experienced three other explosions causing little damage. A survey in 1951 indicated that the utility sustained a loss as high as 38 percent of the total gas fed into the system and at the time of the explosion the leakage had been reduced to 25 percent. As a result of the fire the utility abandoned the underground system and converted all services to individual bottled gas units. If the utility had followed up its "lost and unaccounted for gas" program by definite action and abandoned its services this explosion perhaps could have been avoided.

3.1.3 Casinghead Gasoline Pipeline Break - Los Angeles, CA - February 27, 1956.

Numerous fires, including 13 in buildings, occurred in an area 3.2 kilometre long and 0.8 kilometre wide after a contractor's mechanical ditch digger broke a sewer line, then a casinghead gasoline transmission line. An estimated 80,000 litres of casinghead gasoline (Reid vapour pressure 410 kPa) under pressure escaped from the transmission line and entered the broken sewer. As casinghead gasoline flowed through the sewer it rapidly vaporised and built up sufficient pressure to force its way through water traps in sewer service connections in buildings.

3.1.4 Gas Leakage into Water Main - Long Beach, CA - December 17, 1959.

Workers were cleaning a new but uncompleted 600mm and 760mm water main which was 4.2 kilometres in length. Access chambers were placed about 300 metres apart and most of the access chamber covers had been installed and back-topped over. When two men, involved in the cleaning operation, did not come out at the end of the shift, two others entered a nearby access chamber and struck a match causing an explosion. Both of these men were seriously burned; one died later in the hospital. The two men for whom they were looking either died from asphyxiation or from an explosion which might have occurred several hours before. One other man died from asphyxiation in a rescue attempt. About 45 minutes after the explosion a fire broke out at another access chamber about 1.1 kilometres from the explosion, apparently caused by fire propagating through the pipe. Fire fighters tested the access chamber with combustible gas indicators and found flammable mixtures at several locations. Fans were used to clear the pipe before the victims were recovered. Apparently natural gas leaked into the pipe from an unknown source.

3.1.5 Gasoline in Sewer - Indianapolis, IN - December 8, 1966.

After heavy rainfall, a 23,000 litre gasoline tank in a service station under construction settled in the trench. The pipe connections broke, allowing about 3,800 litres of gasoline to flow into the storm sewer. Explosions occurred about 2.4 kilometres downstream, rocketing steel access chamber covers high in the air and causing severe damage to the sewer line. The ignition source was not definitely determined but might have been caused by sparking from electrical cables in the sewer chambers.

3.1.6 Electrical Fault - Seattle, WA - August 17, 1968.

Failure of three 13,000 volt power lines in a street vault caused fire and explosions. Power to secondary lines was knocked out, causing complete power failure over a 40-block area. One access chamber cover was blown 12 stories high.

3.1.7 Gas Leakage in Access Chamber - New York, NY - October 15, 1968.

Eight fire fighters and Consolidated Edison workers were injured when a gas explosion and fireball erupted from an access chamber at 42nd Street and Eighth Avenue. Subway and street traffic was snarled, and some buildings were evacuated. The accident was caused by ignition of gas leaking from a 50mm pipe. Although the type of gas was not mentioned, it was probably natural gas. The source of ignition was not definitely established.

3.1.8 Gasoline In Sewer - Joliet, IL - July 2, 1969.

A series of sewer explosions in the downtown business district blew out access chamber covers and many windows. Exposed buildings were also damaged by fires, and large chunks of concrete were torn form the wall of the Illinois Waterway. The explosions were caused by ignition of gasoline vapours in the sewer lines. The source of the gasoline was a storage tank on the premises of a bottling plant. Vandals had broken into the plant and stolen some motor oil and possibly some gasoline. The pump was found running, and it was estimated that 3,200 litres had entered the sewer system through a catch basin in the yard.

3.1.9 Gasoline in Sewer - Nashville, TN - January 18, 1970.

Gasoline that leaked from a bulk storage plant entered the sewer system and, eventually, the sewerage treatment plant, resulting in two explosions. The piping for the roof drain of a floating roof tank, which goes down inside the tank, had frozen and ruptured. Since the valve at the discharge end had been left open, gasoline flowed from that drain valve into the diked area surrounding the tank. The dike drain had also been left open, allowing the gasoline to flow out of the diked area into an open sewer connection nearby. The gasoline eventually made its way to the treatment plant, where the explosions occurred. It was estimated that 175,000 litres of gasoline entered the sewer system.

3.1.10 Flammable Liquids in Sewer - Peabody, MA - May 23, 1972.

Flammable liquids from an unknown source leaked into the sewerage system and ignited from a welding torch at the sewerage treatment building. The ensuing fire completely damaged the building.

3.1.11 Gas in Access Chamber - Savannah, GA - June 1973.

An unknown gas had accumulated in an access chamber and was ignited by a welder's torch. All the circuits in the manhole were de-energised by the explosion and the ensuing fire was extinguished by carbon dioxide extinguishers. 

3.1.12 Flammable Liquids in Sewer - Akron, OH - June 23, 1977.

Liquid chemicals used by a rubber products manufacturer were dumped into the city's sewer system intentionally by unknown persons. The chemicals consisted of 11,400 litres of petroleum naphtha, and smaller amounts of acetone and isopropyl alcohol. Multiple explosions occurred within a six-block area causing damage to two main streets and a playground. The source of ignition was not determined. Fire fighting operations consisted of flushing the sewer system. There were no reported building fires. Damage was estimated at $(US)300,000.

3.1.13 Gasoline Spill into Sewer System - Wapello, IA - March 29, 1979.

In this incident gasoline spilled from a tanker while a delivery was being made to a gasoline service station. It was estimated that approximately 110 litres entered the sewer system. Explosions blew access chamber covers off in a six-block area. Basements of buildings in a 20-block area were checked for fires. Two fires were discovered in basements and quickly extinguished by fire fighters. Fire fighting operations also consisted of opening hydrants in the affected area to flush the sewer system. Property damage to the sewer system and buildings was not reported.
Prentiss & Laughlin (1995) refer to fires and explosions that have occurred more recently in the water/wastewater industry. These incidents are documented by the National Fire Protection Association USA. Details of these incidents are as follows, numbers 3.1.14 - 3.1.19.

3.1.14 Natural gas leak - Massachusetts, October 1975

In this incident a sewer line was under construction. Fire ignited natural gas which had built up due to a gas line leak 15 metres from the incident. An explosion occurred when a worker tried to relight a gas-fuelled lantern that had gone out.

3.1.15 Gasoline in sewer - Michigan, June 1988

An arc from a pump ignited vapours from gasoline that spilled into the sewers from a tank farm. Three major explosions occurred. The first took the wood roof off the brick-walled building of a water pumping station.

3.1.16 Gasoline leak into sewer - Massachusetts, August 1988

Gasoline leaked into the sewer during removal of a 7,600 litre underground tank. Dozens of access chamber covers were blown into the air. Explosions blasted holes in front lawns and ripped up 4.6 metres of asphalt sidewalk and a concrete wall.

3.1.17 Gasoline in sewer - Illinois, January 1989

Gasoline entered a sanitary sewer line and vapours were ignited by workers working on a restaurant roof near a "soil stack." Several access chamber covers were blown into the air.

3.1.18 Methane explosion in a pump house - Illinois, May 1989

Sparks from electrical controls ignited a methane-air mixture. One wall of the pump house was blown out and the pump house was blown off its foundation.

3.1.19 Propane or methane gas explosion in a sewerage treatment plant - July 1989,

New York

Propane gas or methane gas built-up led to an explosion in a sewerage treatment plant. Propane was stored in a tank 4.6 metres from the structure. Pieces of the building were found as far as 46 metres away.
The following incidents occurred in countries other than the U.S.A.

3.1.20 Petrol and hexane in sewer - Guadalajara, Mexico, 22 April 1992

In Guadalajara, Mexico, on 22 April 1992 explosions in the sewerage system killed 190 people and injured 1470 people (Reuters 1992 & Bishop 1992). The explosions started in a stagnant blind end created by a temporary diversion for a new railway transit system with further explosions throughout the sewer network. The flammable atmosphere was caused by a leaking petrol pipeline and hexane from a cooking oil factory which leaked into the system. Methane gas from sewage decomposition may have also been accumulating in the stagnant site. At least nine explosions in the system devastated 20 blocks of Guadalajara's La Reforma neighbourhood. Residents had been complaining for days that the sewers reeked of petrol, but neither the governor of Jalisco State (that includes Guadalajara) nor the mayor took any notice. The Mayor and eight other officials were indicted on charges including negligent homicide. The mayor and officials of the government oil company, Premex, were held for trail without bail.

3.1.21 Solvent in sewer - Rhodes, NSW, 26 September 1986

On 26th September 1986 an explosion occurred in a sewerage pumping station at Rhodes, Sydney. The station pumps sewerage from nearby industries which included a paint factory and a large chemical plant. The blast overturned an 8-tonne concrete slab forming part of the roof and hurled a heavy steel cover from the machinery well about 100 metres away. If the hatch cover had been blown westward, it could have struck a large sphere containing propane gas. Broken bricks and other debris were found 100 metres away from the station. As the blast occurred a large fireball shot into the air (Sydney Water Board 1986).

3.2 Fire and Explosion risks when accessing sewers

Gerardi (1982) mentions that sparks created by removing an access chamber cover could cause an explosion if a flammable gas were present. He suggests when testing an access chamber that the presence of flammable gases should be checked first, then asphyxiating and toxic gases and then oxygen deficiencies. To prevent sparks from the removal of a cover he suggests that gas readings be taken by inserting the meter probe of the gas meter through the "pick holes" of the cover. Where "pick holes" are not present, Gerardi recommends that the cover be carefully removed without causing sparks. After the cover has been removed, further testing for dangerous gases should be made in the lower levels of the access chamber since carbon dioxide, hydrogen sulphide and gasoline vapours may have accumulated in the lower levels of the manhole, and their presence would not have been detected in the upper portion of the chamber. Gerardi does not give any details on how to remove a cover "carefully". He mentions potential sources of ignition such as: motor from ventilation equipment, smoking and radio equipment. Appendices C to H show designs for a number of different types of access chamber covers used by the Provider. Appendix G shows a cover with a hole which goes right through the centre. A plastic plug is used to seal the hole. Gas testing with this type of cover is easy because a sampling probe can be inserted through the hole after the plastic plug has been pushed out (a new plug can be inserted when the cover is removed). Unfortunately, in the sewerage system operated by the Provider very few covers have holes passing right through to enable gas sampling by means of a probe.
For a fire or explosion to occur three conditions need to be present:

      1. a source of fuel
      2. sufficient oxygen
      3. an ignition source
Ignition sources fall into the following general categories.
The question, of whether the operation of removing an access chamber cover poses an ignition risk, was researched.
In this operation potential sources of ignition are as follows:

3.2.1 Ignition by friction and impact

Frictional heating occurs when mechanical energy is expended in doing work against friction; the energy dissipated is transformed into heat at the rubbing surfaces. During the process of lifting an access cover mechanical energy is expended against gravity and against friction between the cover and its surround. Bowden and Tabor (1954) and Ward et al. (1990) have shown that when bodies slide (or rub) on each other they touch at very few points. Sliding friction leads to a series of small hot-spots whose temperature does not usually exceed that of the lower of the melting points of the two bodies.
It has been found possible to ignite methane-air by friction between two materials. All the other well-known flammable gases and vapours, with the possible exception of ammonia, are capable of being ignited by the same means (Powell 1969).
Frictional impact (surfaces impact at other than 900) and rubbing may ignite gases and vapours either by producing a hot surface or surfaces or by projecting hot and possibly reactive particles into the surrounding atmosphere. See section 3.23 for a discussion on ignition by hot surfaces.
The risk of ignition from dropping an access chamber cover was researched. Powell and Quince (1972) have investigated the risk of ignition of flammable gases and vapours from the impact between two bodies. They emphasis that it is not immediately obvious that friction during impact plays an important part in the production of surfaces hot enough (usually greater than 1100oC) to cause ignition. Given the physical properties of two bodies undergoing impact they have given an equation that can be used to predict the heat energy dissipated at the surfaces in contact and the resulting surface temperatures.
The total energy dissipated during impact E = Wf + Wr

where E = kinetic energy of falling object

Wf = energy dissipated by friction

Wr = energy dissipated by deformation of the object

Wf = 0.4 (krc)1/2qmaxAT1/2

where k = thermal conductivity

r = density

c = specific heat

A = Area of impact between the two surfaces

T = Impact time

qmax = maximum temperature, at the point of impact, resulting from heating of the surfaces undergoing friction

To use this equation to determine the maximum temperature obtained, the variables Wf, A and T have to be determined by experimentation. Wf and Wr are determined by measurements of the translational and rotational velocities of the falling object before and after impact.
Powell and Quince (1972) conducted experiments where weights were dropped onto surfaces at various angles. Graph I shows energy loss due to friction impulse and normal impulse as a function of angle of inclination between the two surfaces. The temperatures of the surfaces at the point of contact are also shown.

Graph 1 (Source Information Not Available)

Energy dissipated by normal & frictional impulse, & temperature measurements

(Source: Powell and Quince, 1972, pp.1546)

 The graph shows that when the angle was small most of the energy was lost as normal impulse (deformation of the surfaces) with low resultant temperature at the point of impact. As the angle was increased less energy was lost as normal impulse and more as frictional impulse (heating of the surfaces). Loss of energy as frictional impulse peaked at an angle of about 50o but maximum temperature at the point of impact occurred with an angle of about 60o.
In principle the equation given by Powell and Quince can be used to predict the resulting temperature from two impacting bodies provided the situation under which impact occurs is known and provided accurate measurements are made. The equation could be used to predict temperatures generated from dropping an access chamber cover onto the edge of its surround. However, much work would be required to assess all the different types of access chamber covers and surrounds available in the Provider's area of operation. This work in itself would be a major research project.
Appendices C to H show different designs for various access chamber covers. Since covers are tapered at various angles there is a possibility, if one is dropped, for impact to occur at an angle with the surround, with the possibility of frictional heating at the point of contact and/or generation of a spark.
In their experiments, Powell and Quince obtained temperatures of about 430oC when two objects collided at an angle of 61o with an energy loss (from the falling object) of about 3.7 Joules. Access covers weigh up to 120Kg. A cover weighing this much would only have to fall 3.1 mm to generate 3.7 J of work.

Work = force x distance

= mass x accel due to gravity x distance

3.7 = 120 x 9.8 x d

d = 3.1 mm #9;

If the cover were lifted from one edge only, the raised edge would need to fall only 6.2 mm to generate 3.7 Joules of work. If the angle of impact was virtually normal little rise in temperature would occur.
In a later paper, Powell, Billinge and Cutler (1975) pointed out that during normal impact i.e., when the impacting forces have no tangential component, friction plays little or no part. Where there is penetration of one body by the other, heat produced by any resulting plastic deformation is dissipated throughout the volume of deformed material and the rise in surface temperature is generally small.
One of the most useful journal articles on impact and ignition is by Powell (1986). He has summarised in Table 1 results of impact tests which have ignited or failed to ignite various types of flammable gases/vapours-air mixtures. In addition, the occurrence of ignition or non-ignition is related to impact energy. Table 1 is shown below.

Group

I or N

Impact < 200J

Impact > 200J

Group I #9; methane

I

Ce,Zr,Hr,Ti on hard materials

Hand hammer blows on light metal smears on rust

Mg,Ti,Al, and alloys on rusty steel

Chrome-steel on sandstone and carborundum

Al bronze, CuBe, glass on sandstone

Tool steel, bronze, nickel chrome steel on sandstone

Brass on corroded magnesium anode

Very hard steel on very hard steel

Carbon steel on carbon steel

Rotary impact high carbon steel on high carbon steel

Tungsten carbide, steel on sandstone

Hobnail on sandstone

 

N

Mild steel on sandstone

Tungsten carbide on sandstone

Zn,Cd,Al bronzes on rusty steel

C Steel, CuBe, Brass on rusty steel

Mild Steel on sandstone

Group IA

Propane etc.

I

Carbon Steel on Carbon Steel

Hobnail, tungsten carbide, Al bronze, CuBe Tool steel, Lamp glass on sandstone

Rotary impact carbon steel on carbon steel

Rotary impact carbon steel on mild steel

Mild Steel on Mild Steel (600J)

 

N

Mils steel on rusty mild steel (180J)

Hard steel on hard steel (180J)

Al bronze on rusty steel (180J)

Rotary impact of steel on steel

Group IIB

Ethylene

etc.

I

Bottle glass on sandstone

Tool steel on rusty mild steel

Mild steel on rusty mild steel

Al bronze on clean mild steel

 

 

N

Tool steel, Mild steel on clean mild steel

Brass, CuBe, Al bronze on rusty steel

 

Group IIC

Hydrogen #9; etc.

I

CuBe on clean and rusty mild steel

Al bronze on clean and rusty mild steel

Silver steel on rusty mild steel

CuAl alloy on rusty steel

Mild steel on rusty mild steel

Rotary impact bronzes on steel

CuNi alloy on rusty steel

CuAlNi alloys on rusty steel and hard steel

CuZnAl alloy on rusty steel

 

N

Brass on rusty mild steel

Silver Steel on clean mild steel

Zinc alloys on steel

Table 1: Summary of Impact Tests

(Source: Powell, 1986 pp.424)

Palmer (1960) cited in Powell (1986) gives information on fires that have been caused by impact and friction between objects. In 1956 and 1957, in the UK, the number of fires attributed to "friction sparks" were 338 and 341 respectively. In the two years, 30 fires were ascribed to low energy impacts (such as hand tools). In one incident, town-gas was ignited by the impact of a steel hand tool on brickwork. In another, town-gas was ignited when a heavy metal object was moved against a steel and concrete structure. In three incidents petrol was ignited by friction between metal boot studs and concrete. In another three incidents petrol was ignited by heavy metal objects being moved on steel and concrete structures. All of these incidents were described as low energy impacts.
Billinge (1979) cited in Powell (1986) published a survey of incidents between 1958 and 1978 attributed to impact and friction involving gases, vapours and dusts in industries other than mining. 68% of the gas and vapour ignitions were attributed to impact, and 60% of these impacts ignitions were medium energy impacts (i.e. involving about 14J of energy). Seven ignitions of petrol vapour were caused by low energy impact (i.e. about 10J), including some from impact between steels. Other low energy impacts involved light metals or more easily ignited gases.
In certain situations, particularly roadways, access chamber covers become firmly wedged due to vehicles passing over them. In these situations, a hammer and chisel is used to loosen the covers. Some respondents to the survey said that sparks are generated during the process. It would be reasonable to assume from the data previously presented that sparks may be a potential source of ignition.

 

3.2.2 Ignition by friction with imbedded rock particles

Powell, Billinge and Cutler (1975) also considered the possibility of ignition by piezoelectric effects from mechanical stress on crystalline quartz. They concluded that it is extremely unlikely for all the crystals in a quartzitic sandstone to be similarly oriented to give a net charge when stressed.
A crystal is said to show piezoelectricity if an electric charge is developed on its surface from pressure exerted at the ends of its axis. Only those minerals crystallising in symmetry classes that lack a symmetry centre and thus have polar axes can show this property. Quartz is probably the most important piezoelectric mineral, for an extremely slight pressure parallel to an "electric axis" (any of the three a axes) can be detected by the electric charge set up (Hurlbut 1952).
Rae (1964) discounted the influence of the piezoelectric effect after recording similar times for ignition from friction when using sandstone sliders on sandstone and when using non-piezoelectric fused quartz sliders on sandstone. Kocherga (1969) maintained that in the impact of steel or bronze on sandstone, the temperature of heating of the impact point was of little importance in the ignition of methane-air mixtures and that the piezoelectric properties of the quartz crystals were crucial. Work done by the Safety in Mines Research Establishment shows that the typical delay times of about 25ms, measured between the moment of brightest heating during intermittent cutting of sandstone by a coal-cutter pick, and the first signs of ignition of 7% methane-air, are not consistent with ignition with electrical discharges, where a much shorter delay would be expected (Powell, Billinge and Cutler 1975). On the balance of information the piezoelectric effect is not the only factor in the ignition of flammable gases/vapours-air mixtures.
Friction occurring in the presence of quartzitic rocks is known to cause ignition of methane-air mixtures. Quartz has this ability because of its relatively high melting point (Rae 1964). This may be an important factor to consider when removing access chamber covers because there may be sandstone in the gap between the cover and its surrounds. The Provider is in an area where sandstone is present in the soil (Jones 1968).
When the Technical Information Sheet (Appendix A) was written, the possibility of ignition (during lifting) due to friction between the access cover and its surround was considered. At the time, because insufficient information was available, it was recommended that water be used to wet the seal. This was based on the observation of a number of people (Kocherga 1969, Powell, Billinge & Cutler 1975, Powell & Billinge 1975) who found that water was beneficial in suppressing ignition during the cutting of coal.
However, Powell (1986) mentions that in rubbing and cutting experiments in which speed was a variable, speed was shown to be very important in its ability to cause ignition. Lewis, Smith and Powell (1983), cited in Powell (1986), investigated the risk of ignition from cutting sandstone with machine picks in coal mines. They showed that the hazard decreased linearly with speed from 4.5 m/s to 1.5 m/s and then changed abruptly in the region of 1 m/s to 0.5 m/s below which the hazard decreased much faster with speed. Since, during the process of cutting there is always some friction present, the work of Lewis, Smith and Powell suggests that the risk of ignition during the process of lifting an access chamber cover is low. In addition, data from Powell (1986) supports this view. Friction of mild steel on mild steel at 10m/s requires energy to be dissipated at a rate of about 3700W to ignite methane-air mixtures; and 3% P cast iron rubbing on mild steel at speeds of up to 10m/s will not ignite methane-air mixtures. Cast iron used in access chamber covers conforms to various Australian Standards which specify tensile strength and hardness, but do not give the composition of elements such as phosphorus.
 

3.2.3 Production of hot surfaces and sparks by friction and the relative ease of ignition of flammable gases and vapours

As mentioned previously, friction may ignite gases and vapours by producing a hot surface(s) or by projecting hot and possibly reactive particles into the surrounding atmosphere.
Rae et al. (1964) investigated the size and temperature of a hot surface required to ignite a 7% methane-air mixture. The hot surface was an electrically heated square of alumina set flush with the wall of an explosion box. They found that with a temperature of 1250oC a surface of 6 mm square is needed to ignite the gas mixture.; a temperature greater than 1600oC was required if the hot surface was less than 2.5 mm square; and a temperature in excess of 1100oC was required, even with a surface as large as 20mm square.
The ability of hot particles to ignite various flammable gas-air mixtures was investigated by Silver (1937), cited in Powell (1969). He found that a platinum sphere at 1200oC had to be at least 6.5 mm in diameter to ignite 8% methane-air. Coal gas-air, pentane-air, and hydrogen-air were ignited more easily; quartz and platinum sphere only 1 mm in diameter, at a temperature of 1200oC, were found to cause ignition in these gas mixtures.
Hot particles, such as cerium alloys, aluminium, magnesium and titanium, oxidise, with an increase in temperature, as they pass through the atmosphere. Consequently smaller masses of these materials will ignite flammable gas-air mixtures (Powell 1969, 1986). However, it is not now thought that sparks, which are small incandescent or reacting particles torn from the rubbing or impacting surfaces, are the main sources of ignition unless there are circumstances which make the sparks particularly reactive. It is known to be a relatively difficult matter to ignite gases or vapours by hot projected material (Powell 1969). However, since access covers are made from cast iron, or cast iron and concrete, these light metals and flints are not relevant to this area of research.
Mixtures of flammable gases/vapours with air are flammable within certain ranges; the lower limit of flammability is known as the lower explosive limit and the upper limit of flammability is known as the upper explosive limit. However, with ignition by steel sparks it has been found that the lower limit is unchanged, but the upper limit for ignition is often significantly lowered [Voigtsberger (1955) cited in Powell (1986)]. This in effect is equivalent to reducing the flammability of the gas or vapour/air mixture.
Although gases and vapours are not classified into groups by means of ignition by frictionally heated surfaces, sufficient is known to suggest that the groupings used for electrical apparatus in explosive atmospheres can be used as a reasonably accurate guide to the ease of ignition by friction (Powell 1969, 1984 & 1986). Table 2 shows electrical grouping for a number of flammable gases and vapours. Gases/vapours-air mixtures in group IIC are more easily ignited than gases/vapours-air mixtures in group I (the order of ease of ignition is IIC>IIB>IIA>I). From the table it can be seen that hydrogen-air is more easily ignited than ethylene-air and ethylene-air is more easily ignited than propane-air, whilst propane-air is more easily ignited than methane-air.

 

Electrical Division Group

Gas or Vapour

Auto-Ignition Temperature oC

IIC

Hydrogen

Acetylene

Carbon disulphide

585

305

102

IIB

Ethylene

Towns gas

Carbon monoxide

425

449 - 499

570

IIA

Propane

Butane

450

365

I

Methane

538

Table 2: Electrical Grouping for Flammable Gases and Vapours

 

Another characteristic of the ignition of gases/vapours-air mixtures by friction is that it is found that the most easily ignited mixture, is the mixture close to its lower limit of flammability. The same appears to apply for ignition by hot surfaces. These observations apply to methane, although there is evidence that pentane and, possibly, propane depart from this rule. However, one group of researchers did not find that weak mixtures were the most easily ignited when the ignition source was a burning aluminium particle (Powell 1984).
The most easily ignitable mixtures of methane-air found were 7.24% and 5.6%. At least one theoretical consideration has been able to account for lean methane-air mixtures and rich propane-air mixtures being the most easily ignited of those mixtures by hot inert particles (Powell 1984). Strange and MacKenzie-Wood (1985) give 7.5% methane-air as the most easily ignited mixture. However, this relates to an electric spark as the source of ignition. For a flame, the most easily ignited methane-air mixture is 10%, and for a heated surface 6%. The authors also point out that in relation to most flammable gases methane has a relatively long lag in ignition, varying from one-third of a second at 9.5% to two seconds at 5-6% and eight seconds at 12-14%, however, they do not mention the type of ignition source.
The observation that lean methane-air mixtures are the most easily ignited is relevant to the sewerage industry because in two incidents where leaks of natural gas into the sewerage system of an Australian Water and Sewerage Provider occurred, concentrations have been just above 100% LEL (5% v/v) and 180% LEL (9% v/v) (verbal communication with Water Board, Sydney). From Powell (1984), the lower of the two gas concentrations may pose the greater risk because it may be more easily ignited.
On a simple flammable gas/oxygen ratio it seems reasonable that a lean flammable gas-air mixture would be more easily ignitable. As the flammable gas level increases the amount of oxygen is reduced.
The ignition temperature of a gas is the minimum temperature at which the gas will ignite and sustain combustion when mixed with air, without initiation of ignition by spark or flame. The ignition is due to chemical reactions initiated by the temperature of the local environment, and may therefore in practice be a result of the temperature of hot surfaces adjacent to the flammable material. A direct result of established ignition temperatures is the limitation of surface temperatures of equipment in flammable hazardous areas (Standards Australia HB13/NEEITC 181-1-1992 1992).

Table 3 shows ignition temperatures and other properties of methane, toluene and natural gas.

Material

Ignition temperature oC

Explosive Limits (v/v)

Relative Vapour Density

 

 

LEL *

UEL +

 

methane

538

5

15

0.6

toluene

480

1.2

7.1

3.1

natural gas

680

5

15

0.61 - 0.63

* LEL = lower explosive limit

+ UEL = upper explosive limit

Table 3: Ignition temperature & other properties of flammable gases and vapours

 

Powell (1984) points out that a review on the subject of ignition temperature showed that "ignition temperature", or more properly "auto-ignition temperature" of a gas or vapour was not a guide as to whether or not a hot surface will cause ignition. Hydrogen-air can be ignited by sparks or hot surfaces much more easily than methane that has a lower "auto-ignition temperature".
The work of Rae (1964) and Silver (1937) mentioned above show that the temperature of hot surfaces and particles to ignite different flammable gases/vapours-air mixtures depends not only on the gas or vapour, but also on the size of the hot particle or surface. Small particles must be at temperatures well above the "auto-ignition temperature" to cause ignition. Larger hot surfaces ignite flammable gases/vapours-air mixtures at lower temperatures. Powell (1986) points out that many of the tests carried out on grinding sparks limit the grinding time to 5 seconds - otherwise ignition is more likely to be caused by the lower temperature, but larger area of the hot surface produced at the point of contact of the material and the grindstone. Similarly, during glancing impacts, although sparks may be produced, larger contact areas at lower temperatures are more likely to cause ignition.
 

3.3 Non-sparking tools

Non-sparking tools are made from alloys of beryllium-copper, aluminium bronze and brass and have been used around flammable gas areas e.g. petroleum installations and refineries for many years in the belief that they provide a greater degree of safety from the risk of incentive sparking than ferrous hand-tools. Whether non-sparking tools provide any significant safety advantage has been questioned by the Department of Scientific and Industrial Research and Fire Offices' Committee (1963). Their conclusions were:
In the presence of more easily ignited gases such as hydrogen, the hazard arising from the impact of steel on steel may be appreciably reduced if "non-sparking" tools are used. However, the greater risk due to impact of metal on rock still remains whether ferrous or non-sparking tools are used.
Barker (1991) mentions the use of non-ferrous tools and the use of non-sparking surfaces in fireworks factories in the UK. Foster (1979) places a mandatory requirement on the use of non-sparking tools when accessing sewers. Unfortunately neither of these authors give more details nor elaborate on the degree of risk reduction from the use of such equipment.
Powell (1969) states that no non-sparking metal should be regarded as safe if there is a possibility of its being struck against sandstone or concrete containing quartzitic material. He states examples of where ignitions of methane-air mixtures have occurred by striking sandstone with a bronze pick. Laboratory experiments with a machine designed to give blows of the kind equivalent to a hand pick, showed that impact of both steel and brass picks on sandstone produced ignitions of methane-air.
The literature search indicates that non-sparking tools cannot guarantee lack of sparks in use and they could give a false sense of security. They are not as robust as steel tools and need replacement or repair on a regular basis. They are also expensive.
 

3.4 Other sources of ignition

Other sources of ignition within sewers also need to be controlled. It is common for electrical equipment such as submersible pumps, flow sensors, CCTV (closed circuit television) and gas detectors to be used in sewers (Instrumentation In The Water Industry 1976, Pilkington & Bridger 1994, New Technology 1994). The publication, Instrumentation In The Water Industry (1976), makes a point that since the composition of the atmosphere in sewers is unknown, the use of certified intrinsically safe equipment to the highest standard of safety and the highest gas classification IIC (hydrogen) should be used.
Electrical equipment used in flammable gas/vapour hazardous zones must have the appropriate certification. A naturally ventilated sewer is considered non-hazardous, however, events can occur e.g. trade waste discharges and natural gas leaks which make the sewer a flammable gas/vapour hazardous zone.
 

3.5 Other factors

The literature search has revealed a number of important issues which were not directly related to the main aims of the research.
Techniques which are used to control flammable materials in sewers can exacerbated the problem e.g. the use of dispersants. The presence of flammable gases/vapours in sewers is controlled by:
Dispersants have been used for flushing fuel spills from streets into sewers and have been added to sewers in which fuel has been spilled. This has been done in the belief that the fire potential is reduced. Fingas, Hughes and Bobra (1989) have found from the results of small-scale models that the use of dispersants will in most cases greatly increase the potential for explosions and their magnitude. Dispersants increase the rate of volatilisation of small fuel molecules and increase the total amount of these released.
The literature search has highlighted the phenomena of surge within a sewer which could affect the movement of toxic and flammable gases in the system (Guo & Song 1991). Flow in a dropshaft is unsteady during storms or when there is a faulty gate control. Under certain conditions, periods of free flow and pressurised flow may coexist and be separated by a moving surge front. The magnitude of a such a surge may build up significantly as it moves upstream, possibly generating severe water hammer pressures when it reached the upstream boundary. Whenever the surge in the main tunnel or the branch tunnel reaches the dropshaft-drift tube system, the system receives a pressure rise at the downstream end. This can result in excess pressure upon and loss of access chamber covers or dropshaft covers in sewer systems. If the dropshaft is ventilated the cover could not be ejected by air pressure alone. Most cover ejections are caused by the impact force of the rising water. The finding is significant if an access cover is removed while there is flammable gas in the system and a surge occurs. Flammable gas will be released through the opening and may encroach into areas where ignition sources are present i.e. motor cars. However, under normal conditions, a distance of 3 m from an access opening is considered a safe distance for ignition sources.
A technique which is used to suppress the risk of flammable gas/vapour-air ignition is the use of water fogs. Some Water Authorities use this (verbal communication with Western Australian Water Authority). Kletz (1993) says that in exceptional circumstances, a person might be allowed to enter a flammable cloud to close a valve and stop a leak which would otherwise spread. If possible, the person should be protected by water spray. However, no-one should be asked to do so.
However, the National Institute of Occupational Safety and Health (NIOSH) (1985) reported that this technique failed in an incident resulting in the loss of a life. A firefighter was killed and 10 others injured by an explosion while attempting to rescue a worker who had fallen into an empty toluene tank. The rescuers were cutting through the side of the tank with a power saw under streams of water fogs when the explosion occurred. The investigation showed that contrary to accepted beliefs, the use of water fogs did not prevent or reduce the risk of igniting explosive atmospheres and that, under certain conditions their use could in fact increase the severity of such fires or explosions. NIOSH did not give details on how the use of water fogs could increase the severity of fires or explosions.

 

4. Ignition Sources

For an explosion to occur it is essential that an ignition source of sufficient energy to initiate flame propagation is present. The probability of explosions occurring can therefore be reduced by removing known sources of ignition.
Devices whose parameters can never exceed any of the values; 1.2V, 0.1A, 20mJ or 25mW
are considered safe for use in an explosive atmosphere providing they are not connected to a device which contains a source of energy which could cause the circuit to exceed these parameters (Standards Australia 2380.1 1989).
A certain minimum ignition energy, which differs from one flammable gas to another, is required for an explosion to occur. If a source of ignition, such as a spark, has an energy below this it cannot cause an explosion. The minimum ignition energy of a flammable gas-air mixture is the minimum energy required to ignite the most easily ignitable mixture of that gas. The minimum ignition energies of flammable gases are typically in the range of 0.019mJ (for hydrogen) to 0.29 mJ (for methane) (Standards Australia HB13/NEEITC 181-1-1992 1992).
 

4.1 Pagers and torches

Pagers and torches should be considered potential ignition sources unless certified intrinsically safe (Standards Australia 1940 1993).
 

4.2 Cellular phones and radios

These should be considered potential ignition sources unless certified intrinsically safe (Standards Australia 1940 1993).
 

4.3 Battery operated watches and hearing aids

Low-energy devices such as hearing aids and totally enclosed objects such as battery-operated watches are not considered to have sufficient energy to ignite flammable gases/vapours-air mixtures (Standards Australia 1940 1993).
 

4.4 Detector tubes

Detector tubes are sometimes used to measure sewer atmospheres when trade waste discharges are suspected e.g. paint solvents. Care is required because with some detector tubes (e.g. hydrogen detector tubes) sufficient heat of reaction may be produced in the tube to ignite the atmosphere if gas is present in the explosive range. An incident is known where flammable gases/vapours infiltrated the sewer and was detected on gas meters calibrated for methane (verbal comm with Sydney Water). Analysis of the sewer atmosphere by GCMS (gas-liquid/mass spectroscopy) found that ethanol was present. A routine check with Drager tubes indicated the presence of hydrogen. Hydrogen gas was confirmed by a laboratory specialising in this type of analysis. It was fortunate that when the tubes were used for monitoring that an explosion did not occur.
Drager specifically warns users of their tubes of the risk of ignition of flammable atmospheres. For example with the Hydrogen 0.2%/a tube, Drager states:
"The indicating layer is heated up by hydrogen concentrations of more than 10 Vol.-%. The air sample must not contain additional flammable substances whose ignition temperature is below 250oC - Danger of explosion", (Drager 1994) and for the Hydrogen 0.5%/a tube, Drager states:
"Do not use in potentially explosive areas. Qualify before use with a combustible gas monitor. When the hydrogen concentration is above 3 Vol.-% the catalysis layer heats up during the measurement with a reddish glow" (Drager 1994).
 

5. Hazardous Zones

Flammable gas hazard areas are divided into zones (Standards Australia HB13/NEEITC 181-1-1992 1992). The word Zone is internationally accepted as indicating the probability of the presence of a flammable, combustible or explodable material, and the extent, dimension, shape of hazard area and volume in which hazardous material can be expected. Class I refers to flammable gases and vapours ( Class II refers to combustible dusts).
Areas can be classified into the above zones either by specific occupancy or by assessment.
In the Provider's area of operation, the sewerage system was originally constructed with natural ventilation and considered non-hazardous. Natural ventilation was normally effected by the installation of vent stacks at intervals. In recent times there has been a tendency to remove or block ventstacks for aesthetic reasons or because of odour problems. The net effect of reducing ventilation is to increase the risk of flammable gases and toxic gases building up in the system.
 

6. Flammable Hazards

6.1 Methane

During bacterial degradation carbohydrates, proteins and fats are all initially converted to organic acids . Under aerobic conditions, these acids are then oxidised to carbon dioxide and water. However, under anaerobic conditions oxygen is not available, and so other oxidants or methods of degradation are used by bacteria. Such degradation usually requires special groups of anaerobic bacteria capable of carrying out the specific reactions. During anaerobic digestion that is operating in a stabilised condition (e.g. inside a well controlled digester), two groups of bacteria work in harmony to accomplish the destruction of organic matter. The saprophytic organisms carry the degradation to the acid stage, and then the methane-forming bacteria complete the conversion into methane and carbon dioxide.
Methane-forming bacteria are ubiquitous in nature, and some are always present in domestic wastewater and sludge derived therefrom. Their population, however, is very small compared with that of the saprophytic bacteria. Raw sewage sludge has a relatively low buffering capacity, and when it is allowed to ferment anaerobically, volatile acids are produced so much faster than the few methane bacteria can consume, that the buffers are soon spent and free acids exist to lower the pH. At pH 6.5, methane bacteria are seriously inhibited but the saprophytic bacteria are not until the pH levels fall to about 5. Under such unbalanced conditions, the concentration of volatile acids continues to increase depending upon the solid content of the sludge. Active methane digestion may never develop in such a mixture unless the sludge is diluted or neutralising agents are added to produce a favourable pH for methane bacteria (Sawyer and McCarty 1978).
Sewers are designed so that there is a fairly continuous flow so that stagnation is reduced to a minimum.
 

6.2 Natural gas

Natural gas occurs naturally and after processing and drying contains essentially methane (95%). Odorants, such as mercaptans are added to natural gas so that leaks can be detected by smell. In the Provider's area of operation, large gas customers are fed from secondary mains with a supply pressure of 1050kPa. Domestic mains operate from a pressure of 210kPa. In the Provider's location there are still low pressure mains which operate at 7 kPa, but they are being upgraded. Natural gas has a high ignition temperature of 6800C and a relative density 0.61 -0.63. Consequently where there is a leak, the gas tends to rise and dissipate into the atmosphere.
The low pressure mains operating at 7 kPa used to be the mains for town gas generated from coal. Due to the age and construction of these older mains, the risk of leaks are higher than from the newer mains. Domestic mains operating at 210 kPa are made from nylon. These mains have the advantage of being able to withstand earth movement such as earthquakes but are easily punctured if people do not take care where they dig.
 

6.3 Trade waste discharges

The more common flammable liquids found in sewers are paint solvents and petrol. The Provider has conducted extensive public awareness campaigns to encourage the public to be more thoughtful of what is disposed through the sewers. This campaign has been supported by conducting chemical waste collection services in towns and suburbs so that the public can dispose of unwanted hazardous materials and not flush them down the sewers. Although the campaign has reduced illegal discharges into the sewers, occurrences still occur, including discharges of flammable solvents.
In addition, hazardous materials sometimes find their way into the sewers via the stormwater channels. Spills of hazardous material which enter stormwater channels can still affect the sewers because in some sections of the Provider's area of operation there are still connections between sewers and stormwater channels.

 

7. Survey

Procedures as specified in the Technical Information Sheet (Appendix A) were introduced during 1994 as a matter of urgency to reduce the risk of explosions in the sewerage system. As part of a total quality management program being implemented by the Provider, an assessment of compliance with the Technical Information Sheet was done. The results of a survey on confined spaces (Appendix B) were used to achieved this.
The Provider is divided into five Regions of operation. Each Region has a Business Unit Manager who is in charge of Waste Water. Each Region is divided into Waste Water Areas. An Area is managed by a Local Waste Water Manager. Larger Regions have more Waste Water Areas. Each Local Waste Water Manager has about 20 - 60 people working for him. In total there were 488 sewerage maintenance people at the time the survey was conducted.
 

7.1 Questionnaire design

The confined space survey had been designed to ascertain whether there were differences between Areas in Regions and between the Regions in implementing certain procedures, including those specified in the Technical Information Sheet. To obtain objective replies, the questionnaire had been structured to find out what difficulties or problems people were having in carrying out the procedures. The questionnaire was designed so that the results could be analysed by non-parametric statistics using the chi square test. In this test the number of expected replies in each cell needs to be at least 5 for validity. The total number of people surveyed was 488.
The confined space questionnaire which was anonymous contained 24 questions. Questions of direct relevance to the research were: 1,2,4,7,8, and 16. Question 17 was used to ascertain which Region and Area people worked in. Questions 22 and 23 were used to look at the demographics of the respondents.
Each questionnaire, with return self-addressed envelope and an accompanying letter, was sent to each sewerage maintenance person assuring them of their confidentiality and informed them of the purpose of the survey. The questionnaires were delivered to the Local Waste Water Managers who distributed the individually addressed letters to their respective staff.
Overall a total of 488 people were sent the questionnaire and 269 replies were received. This gave a percentage reply of 55.1%. This was considered a good response considering the stress that sewerage maintenance people were facing at the time because there was a proposal from the Provider that the work of sewerage maintenance be put out to contract which in effect meant that future job security for the current workers was uncertain. Polit and Hungler (1987) suggest that response rates of 30% are quite common for questionnaires.
To achieve a higher response rate, a follow up letter was sent to people in Northern reminding them to return the questionnaire. This was sent about two weeks after the questionnaire was despatched. Only one extra reply was received. Because of the little gain, follow up letters were not sent to the other Regions.
 

7.2 Definition of compliance for questions 1,2,4,7,8,& 16


Question 1:

How far do you keep ignition sources away from access chambers (manholes) when removing the covers?

1 metre

3 metre

6 metre

People who replied that they kept ignition sources 3 metres or more away were considered to comply.


Question 2:

Do you wet the seal of access chamber covers (manhole covers) before lifting?

always

sometimes

never

People were given three options for answering the question: "always", "sometimes" and "never". Compliance was considered to occur only if the reply was "always".


Question 4:

How do you lift/remove an access chamber cover (manhole cover)?

"Gatic" lifter

manhole key

"Liftrite" lifter

Shovel

Lifting chisel

Photographs 1, 2, 3, 5, 6, and 7 show different lifting devices used in the Provider's area of operation. Photograph 4 shows a hammer and chisel being used to loosen a cover; the winch, attached to the rear of the vehicle is used to lift the cover. At the time that the Technical Information Sheet was written the use of a winch was not considered because few people used it.
The "Technical Information Sheet" required the use of a lifting device which maximised the distance of a person from the cover. Only the "Gatic" lifter and the "Liftrite" lifter were considered to comply with this requirement. People who replied to using only these devices were considered to comply.


Question 7:

If you do a gas test, at what stage do you test?

When access chamber cover (manhole cover) is partly lifted (e.g. 50 mm)

When access chamber cover (manhole cover) is completely removed

I don't gas test

For compliance, the first reply was required. If a person replied that they did not gas test and said they were not a gas tester, the reply was not counted as non-compliance. However, a blank reply was counted as non-compliance. The replies to Question 14 were used to help work out some of the replies. For example, if a person gave no reply, it could be because they were not gas testers or that they don't gas test at all. The reply to Question 14 (if answered) would indicate whether a person was a gas tester or not.


Question 8:

If your answer to Question 7 is the first one, where do you place the gas meter on a windy day?

Upwind of access chamber cover (manhole cover)

Downwind of access chamber cover (manhole cover)

It does not matter

For compliance the second reply was required.


Question 16:

Have you seen the Technical Information Sheet "Precautions when removing access covers in the possible presence of flammable gases/vapours"?

This question was used to determine who received the Technical Information Sheet.
 


8. Results of Survey

8.1 Results of survey - Northern Region

The first Region surveyed was Northern. This Region has one Business Unit Manager to whom three Local Waste Water Managers report. The Region is divided into three Areas referred to as Area C, Area H and Area W. The number of people in each of these Areas and the numbers who replied to the questionnaire are summarised below.

Northern Region

Area C

Area H

Area W

Number of people in each area

54

22

37

Number of people from each area replying to questionnaire

20

13

14

Percentage reply

37%

59%

38%


Reply to question 1:

How far do you keep ignition sources away from access chambers (manholes) when removing the covers?

1 metre

3 metre

6 metre

Northern Region

Area C

Area H

Area W

Number of people who keep ignition sources at least 3m away

20

13

13

Number of people who do not keep ignition sources 3m away

0

0

0

Number of people who did not reply to this question (*)

0

0

1

Percentage compliance

100%

100%

92%

(*) Counted as non-compliance.


Reply to question 2:

Do you wet the seal of access chamber covers (manhole covers) before lifting?

always

sometimes

never

Northern Region

Area C

Area H

Area W

Number of people who always wet the seal

2

1

1

Number of people who sometimes wet the seal

15

11

6

Number of people who never wet the seal

3

1

7

Number of people who did not reply to this question

0

0

0

Percentage compliance

10%

8%

7%

 

The chi square test was not done because the number of people expected to comply for Areas C, H and W were 1.70, 1.11 and 1.19 respectively. For a chi square test to be reasonably valid none of the expected replies should be less than 5 (Larsen 1975). Page 61 shows the chi square test detailed.


Reply to question 4:

How do you lift/remove an access chamber cover (manhole cover)?

"Gatic" lifter

manhole key

"Liftrite" lifter

Shovel

Lifting chisel

Northern Region

Area C

Area H

Area W

Number of people who only use "Gatic" lifters and/or "Liftrite" lifters

3

3

1

Number of people who do not always use "Gatic" lifters and/or "Liftrite" lifters

17

10

13

Number of people who did not reply to this question

0

0

0

Percentage compliance

15%

23%

7%

 

The chi square test was not done because the number of people expected to comply for Areas C, H and W were 2.98, 1.94 and 2.09 respectively.


Reply to question 7:

If you do a gas test, at what stage do you test?

When access chamber cover (manhole cover) is partly lifted (e.g. 50 mm)

When access chamber cover (manhole cover) is completely removed

I don't gas test

Northern Region

Area C

Area H

Area W

Number of people who test when the cover is partly lifted (e.g. 50mm)

2

1

1

Number of people who test when the cover is completely removed

17

10

12

Number of people who replied "I don't gas test"

1(+)

1(#)

0

Number of people who did not reply to this question

0

1 (##)

1 (*)

Percentage compliance

11%

8%

7%

(+) This person replied in Question 14 that he was not a gas tester: reply not counted as non-compliance.

(#) This person said he was not a gas tester and gave the same reply in Question 14: reply not counted as non- compliance.

(##) This person indicated in Question 14 that he was a gas tester: reply counted as non-compliance.

(*) This person indicated in Question 14 that he was a gas tester: reply counted as non-compliance.

The chi square test was not done because the number of people expected to comply for Areas C, H and W were 1.69, 1.07 and 1.24 respectively.


Reply to question 8:

If your answer to Question 7 is the first one, where do you place the gas meter on a windy day?

Upwind of access chamber cover (manhole cover)

Downwind of access chamber cover (manhole cover)

It does not matter

Northern Region

Area C

Area H

Area W

Number of people who gas test when the cover is partly lifted (e.g. 50mm) and who place gas meter upwind on a windy day

0

0

0

Number of people who gas test when the cover is partly lifted (e.g. 50mm) and who place gas meter downwind on a windy day

2

1

1

Number of people who gas test when the cover is partly lifted (e.g. 50mm) and who replied that it doesn't matter where the gas meter is placed

0

0

0

Percentage compliance

10%

8%

7%

 


Reply to question 16:

Have you seen the Technical Information Sheet "Precautions when removing access covers in the possible presence of flammable gases/vapours"?

Northern Region

Area C

Area H

Area W

Number of people who have seen the Tech Info Sheet

6

2

5

Number of people who have not seen the Tech Info Sheet

12

11

9

Number of people who did not

reply to this question (+)

2

0

0

Percentage "reached"

30%

15%

36%

(+) Counted as non-compliance.

The chi square test was not done because the number of replies for people expected to have seen the Technical Information Sheet for Areas H and W were 3.60 and 3.87 respectively.
 


8.2 Results of survey - North Western Region

This Region has one Business Unit Manager to whom three Local Waste Water Managers report. The Region is divided into three Areas referred to as Area P, Area B/H and Area BM/P.

North Western Region

Area P

Area B/H

Area BM/P

Number of people

36

39

29

Number replying to questionnaire

12

16

20

Percentage reply

33%

41%

69%

 


Reply to question 1:

How far do you keep ignition sources away from access chambers (manholes) when removing the covers?

1 metre

3 metre

6 metre

North Western Region

Area P

Area B/H

Area BM/P

Number of people who keep ignition sources at least 3m away

11

16

20

Number of people who do not keep ignition sources 3m away

1

0

0

Number of people who did not reply to this question

0

0

0

Percentage compliance

92%

100%

100%

 


Reply to question 2:

Do you wet the seal of access chambers covers (manhole covers) before lifting?

always

sometimes

never

North Western Region

Area P

Area B/H

Area BM/P

Number of people who always wet the seal

4

5

3

Number of people who sometimes wet the seal

7

10

14

Number of people who never wet the seal

1

1

3

Number of people who did not reply to this question

0

0

0

Percentage compliance

33%

31%

15%

 
The chi square test was not done because the number of expected replies for people who always wet the seal for Areas P and B/H were 3 and 4 respectively.
 


Reply to question 4:

How do you lift/remove an access chamber cover (manhole cover)?

"Gatic" lifter

manhole key

"Liftrite" lifter

Shovel

Lifting chisel

North Western Region

Area P

Area B/H

Area BM/P

Number of people who only use "Gatic" lifters and/or "Liftrite" lifters

0

1

0

Number of people who do not always use "Gatic" lifters and/or "Liftrite" lifters

12

15

20

Number of people who did not reply to this question

0

0

0

Percentage compliance

0%

6%

0%

 


Reply to question 7:

If you do a gas test, at what stage do you test?

When access chamber cover (manhole cover) is partly lifted (e.g. 50 mm)

When access chamber cover (manhole cover) is completely removed

I don't gas test

North Western Region

Area P

Area B/H

Area BM/P

Number of people who test when the cover is partly lifted (e.g. 50mm)

8

12

16

Number of people who test when the cover is completely removed

4

3

4

Number of people who replied "I don't gas test"

0

0

0

Number of people who did not reply to this question

0

1 (#)

0

Percentage compliance

67%

75%

80%

(#) Counted as non-compliance because reply to Question 14 indicated that the person was a gas tester.

The chi square test was not done on the three Areas because the number of expected non-compliance replies were 3 and 4 for Areas P and B/H respectively.
 


Reply to question 8:

If your answer to Question 7 is the first one, where do you place the gas meter on a windy day?

Upwind of access chamber cover (manhole cover)

Downwind of access chamber cover (manhole cover)

It does not matter

North Western Region

Area P

Area B/H

Area BM/P

Number of people who gas test when the cover is partly lifted (e.g. 50mm) and who place the gas meter upwind on a windy day

3

2

2

Number of people who gas test when the cover is partly lifted (e.g. 50mm) and who place the gas meter downwind on a windy day

5

10

13

Number of people who gas test when the cover is partly lifted (e.g. 50mm) and who replied that it doesn't matter where the gas meter is placed

0

0

1

Percentage compliance

42%

63%

65%

 
Of the people who followed the Technical Information Sheet by gas testing when the cover was partly lifted, some did not place the gas meter downwind as required. The expected non-compliance replies for Areas P, B/H and BM/P were 1.78, 2.67 and 3.56 respectively. These numbers were too low to do a chi square test.
 


Reply to question 16:

Have you seen the Technical Information Sheet "Precautions when removing access covers in the possible presence of flammable gases/vapours"?

North Western Region

Area P

Area B/H

Area BM/P

Number of people who have seen the Tech Info Sheet

8

9

9

Number of people who have not seen the Tech Info Sheet

2

6

10

Number of people who did not reply to this question (*)

2

1

1

Percentage "reached"

67%

56%

45%

(*) Counted as non-compliance.

The number of people expected to have seen, and not seen the Technical Information Sheet was greater than 5 for each Area. A chi square test was done to see if there was a significant difference between the Areas with regards to the distribution of the Technical Information Sheet.

Statistical Analysis

Explanation of the chi square test

Two observations have been recorded: (1) Area and (2) frequency of people receiving the Technical Information Sheet. If the variables are independent, the true proportions of people (pp) in Area P who have received the Tech Info Sheet should be equal to the true proportions of people (pB/H) in Area B/H who have received the Tech Info Sheet, which should also equal the true proportions of people (pBM/P) in Area BM/P who have received the Tech Info Sheet.
The null hypothesis can be expressed as:

Ho pP = pB/H = pBM/P

If Ho is true a pooled estimate of the true proportions of people (from Areas P, B/H and BM/P) receiving the Tech Info Sheet would be (8+9+9)/48 = 54.2%. Applying this proportion to the total number in the first column of the table, it follows that we should "expect" (26/48)x12 = 6.5 people to have received the Tech Info Sheet. Similarly:
-The expected number of people from Area B/H to have received the Tech Info Sheet is
(26/48)x16 = 8.67
-The expected number of people from Area BM/P to have received the Tech Info Sheet is
(26/48)x20 = 10.83
-The expected number of people from Area P not receiving the Tech Info Sheet is (22/48)x12 = 5.5
-The expected number of people from Area B/H not receiving the Tech Info Sheet is (22/48)x16 = 7.33
-The expected number of people from Area BM/P not receiving the Tech Info Sheet is (22/48)x20 = 9.17

Area P

Area B/H

Area BM/P

Seen the Tech Info Sheet

8 (observed)

6.5 (expected)

9 (observed)

8.67 (expected)

9 (observed)

10.83 (expected)

Not seen the Tech Info Sheet (*)

4 (observed)

5.5 (expected)

7 (observed)

7.33 (expected)

11 (observed)

9.17 (expected)

(*) People who did not reply are included here.

Since the expected frequencies were computed under the assumption that Ho was true, the null hypothesis is rejected if the discrepancies between the observed and expected frequencies are too large. The statistic used to measure the magnitude of those discrepancies is

c2 = (obs - exp)2

---------------

exp

where obs and exp refer to the observed and expected frequencies in a given category (or cell). The sampling distribution is described by a chi square curve with (c-1)(r-1) degrees of freedom, where c = number of columns and r = number of rows. For a chi square test, none of the expected frequencies should be less than 5.

c2 = (0-E)2

--------

E

= (8-6.5)2 + (9-8.67)2 + (9-10.83)2 + (4-5.5)2 + (7-7.33)2 + (11-9.17)2

---------- ----------- ------------ ----------- ---------- ------------

6.5 8.67 10.83 5.5 7.33 9.17

= 1.46

There is no significant difference at p = 0.05 level of significance between the three Areas in the number of people who have seen the Technical Information Sheet.
 


8.3 Results of survey - Central Region

This Region has one Business Unit Manager to whom three Local Waste Water Managers report. The Region is divided into three Areas referred to as Area St.G, Area E and Area C/E.

Central Region

Area St. G

Area E

Area C/E

Number of people

46

38

61

Number replying to questionnaire

35

28

47

Percentage reply

76%

74%

77%

 


Reply to question 1:

How far do you keep ignition sources away from access chambers (manholes) when removing the covers?

1 metre

3 metre

6 metre

Central Region

Area St. G

Area E

Area C/E

Number of people who keep ignition sources at least 3m away

33

27

44

Number of people who do not keep ignition sources 3m away

2

0

2

Number of people who did not reply to this question

0

1

1 (*)

Percentage compliance

94%

96%

94%

(*) Page one of this person's replies was missing: counted as non-compliance.

 


Reply to question 2:

Do you wet the seal of access chambers covers (manhole covers) before lifting?

always

sometimes

never

Central Region

Area St. G

Area E

Area C/E

Number of people who always wet the seal

6

4

13

Number of people who sometimes wet the seal

29

23

33

Number of people who never wet the seal

0

0

0

Number of people who did not reply to this question (+)

0

1

1 (*)

Percentage compliance

17%

14%

28%

(+) Counted as non-compliance.

(*) Page one of this person's replies was missing: counted as non-compliance.

Statistical analysis

The expected replies for the three Areas were greater than 5 for each cell. A chi square
test was done to see if there was a significant difference between the three Areas.

Area St.G

Area E

Area C/E

Numbers who wet the seal

6 (observed)

7.32 (expected)

4 (observed)

5.85 (expected)

13 (observed)

9.83 (expected)

Numbers who do not wet the seal (*)

29 (observed)

27.68 (expected)

24 (observed)

22.15 (expected)

34 (observed)

37.17 (expected)

(*) People who did not reply are included here.

c2 = (6-7.32)2 + (4-5.85)2 + (13-9.83)2

----------- ---------- ------------

7.32 5.85 9.83

+ (29-27.68)2 + (24-22.15)2 + (34-37.17)2

------------- -------------- --------------

27.68 22.15 37.17

= 2.33

There is no significant difference at p = 0.05 level of significance between the three Areas in the number of people who always wet the seal.
 


Reply to question 4:

How do you lift/remove an access chamber cover (manhole cover)?

"Gatic" lifter

manhole key

"Liftrite" lifter

Shovel

Lifting chisel

Central Region

Area St. G

Area E

Area C/E

Number of people who only use "Gatic" lifters and/or "Liftrite" lifters

1

2

0

Number of people who do not always use "Gatic" lifters and/or "Liftrite" lifters

34

26

47

Number of people who did not reply to this question

0

0

0

Percentage compliance

3%

7%

0%

 


Reply to question 7:

If you do a gas test, at what stage do you test?

When access chamber cover (manhole cover) is partly lifted (e.g. 50 mm)

When access chamber cover (manhole cover) is completely removed

I don't gas test

Central Region

Area St. G

Area E

Area C/E

Number of people who test when the cover is partly lifted (e.g. 50mm)

3

6

9

Number of people who test when the cover is completely removed

26

18

27

Number of people who replied "I don't gas test"

5 (+)

3 (#)

9 (*)

Number of people who did not reply to this question

1 (++)

1 (##)

2 (**)

Percentage compliance

9%

24%

24%

(+) Two people replied in Question 14 that they were not gas testers: replies were not counted as non-compliance. Three persons' replies to Question 14 indicated that they were gas testers: replies were counted as non-compliance.

(++) One person did not reply but indicated in Question 14 that he was a gas tester; counted as non-compliance.

(#) Two people replied in Question 14 that they were not gas testers: replies were not counted as non-compliance. One person's reply to Question 14 indicated that he was a gas tester: counted as non-compliance.

(##) This person replied in Question 14 that he was not a gas tester: reply was not counted as non-compliance.

(*) Seven people replied in Question 14 that they were not gas testers: replies were not counted as non-compliance. Two persons' replies indicated that they were gas testers: replies were counted as non-compliance.

(**) One person did not reply to Question 14 also. Since there was an opportunity in Question 14 to state whether a person was a gas tester or not, this reply was counted as non-compliance. The other person replied in Question 14 that he was not a gas tester: reply was not counted as non-compliance.

 

Statistical analysis

Assuming that the three Areas were independent gave all cells expected frequencies greater than 5 except for Area E which gave a frequency of 4.64. This was considered not too far off the recommended expected value of 5 to do a chi square test.

Area St.G

Area E

Area C/E

People who gas test when the cover is partly lifted

3 (observed)

6.12 (expected)

6 (observed)

4.64 (expected)(*)

9 (observed)

7.24 (expected)

People who do not gas test when the cover is partly lifted

30 (observed)

26.88 (expected)

19 (observed)

20.36 (expected)

30 (observed)

31.76 (expected)

(*) Larsen (1975, p.252) mentions that for an r x c table, none, or at the most, very few of the expected frequencies should be less than 5. Since this cell was only slightly less than 5, the chi square test was still considered to be valid.

c2 = (3-6.12)2 + (6-4.64)2 + (9-7.24)2 + (30-26.88)2 + (19-20.36)2 + (30-31.76)2

----------- ---------- ----------- -------------- -------------- --------------

6.12 4.64 7.24 26.88 20.36 31.76

= 2.96

There is no significant difference at p = 0.05 level of significance between the three Areas in the number of people who gas test when the cover is partly lifted.
 


Reply to question 8:

If your answer to Question 7 is the first one, where do you place the gas meter on a windy day?

Upwind of access chamber cover (manhole cover)

Downwind of access chamber cover (manhole cover)

It does not matter

Central Region

Area St.G

Area E

Area C/E

Number of people who gas test when the cover is partly lifted (e.g. 50mm) and who place the gas meter upwind on a windy day

0

1

1

Number of people who gas test when the cover is partly lifted (e.g. 50mm) and who place the gas meter downwind on a windy day

3

5

7

Number of people who gas test when the cover is partly lifted (e.g. 50 mm) and who replied that it doesn't matter where the gas meter is placed

0

0

1

Percentage compliance

9%

18%

15%

 


Reply to question 16:

Have you seen the Technical Information Sheet "Precautions when removing access covers in the possible presence of flammable gases/vapours"?

Central Region

Area St. G

Area E

Area C/E

Number of people who have seen the Tech Info Sheet

11

17

26

Number of people who have not seen the Tech Info Sheet

23

11

20

Number of people who did not reply to this question (+)

1

0

1

Percentage "reached"

31%

61%

55%

(+) Counted as non-compliance.

Statistical analysis

All expected replies were greater than 5

Area St.G

Area E

Area C/E

Numbers who have seen the Tech Info Sheet

11 (observed)

17.18 (expected)

17 (observed)

13.75 (expected)

26 (observed)

23.07 (expected)

Numbers who have not seen the Tech Info Sheet

24 (observed)

17.82 (expected)

11 (observed)

14.25 (expected)

21 (observed)

23.93 (expected)

c2 = (11-17.18)2 + (17-13.75)2 + (26-23.07)2 + (24-17.82)2 + (11-14.25)2 + (21-23.93)2

----------- ----------- ----------- ----------- ----------- -----------

17.18 13.75 23.07 17.82 14.25 23.93

= 6.61

Chi square tables (Larsen 1975, p.304) show that for 2 degrees of freedom, if c2 = (O - E)2/E is greater than 5.88 (p = 0.05 level of significance) the null hypothesis is rejected. That is, there is a significant difference (at p = 0.05) with respect to the distribution of the Technical Information Sheet between the Areas.


8.4 Results of survey - South Western Region

This Region has one Business Unit Manager to whom four Local Waste Water Managers report. The Region is divided into four Areas referred to as Area B, Area L, Area M and Area S.

South Western Region

Area B

Area L

Area M

Area S

Number of people

19

25

19

26

Number replying to questionnaire

12

10

16

8

Percentage reply

63%

40%

84%

31%

 


Reply to question 1:

How far do you keep ignition sources away from access chambers (manholes) when removing the covers?

1 metre

3 metre

6 metre

South Western Region

Area B

Area L

Area M

Area S

Number of people who keep ignition sources at least 3m away

12

10

15

8

Number of people who do not keep ignition sources 3m away

0

0

1

0

Number of people who did not reply to this question

0

0

0

0

Percentage compliance

100%

100%

94%

100%

 


Reply to question 2:

Do you wet the seal of access chambers covers (manhole covers) before lifting?

always

sometimes

never

South Western Region

Area B

Area L

Area M

Area S

Number of people who always wet the seal

3

0

4

0

Number of people who sometimes wet the seal

9

7

8

2

Number of people who never wet the seal

0

3

4

6

Number of people who did not reply to this question

0

0

0

0

Percentage compliance

25%

0%

25%

0%

 
The chi square test was not done because the number of expected replies for compliance were all below 5 for the four Areas i.e. 1.83, 1.52, 2.43 and 1.22 for Areas B, L, M and S respectively. The chi square test was also not done on Areas B and M because the expected replies for compliance were 3 and 4 respectively.
 


Reply to question 4:

How do you lift/remove an access chamber cover (manhole cover)?

"Gatic" lifter

manhole key

"Liftrite" lifter

Shovel

Lifting chisel

South Western Region

Area B

Area L

Area M

Area S

Number of people who only use "Gatic" lifters and/or "Liftrite" lifters

1

0

4

0

Number of people who do not always use "Gatic" lifters and/or "Liftrite" lifters

11

9

12

8

Number of people who did not reply to this question (*)

0

1

0

0

Percentage compliance

8%

0%

25%

0%

(*) Counted as non-compliance.

 
The chi square test was not done because the number of people expected to comply for Areas B, L, M and S were 1.30, 1.09, 1.74 and 0.87 respectively.
 


Reply to question 7:

If you do a gas test, at what stage do you test?

When access chamber cover (manhole cover) is partly lifted (e.g. 50 mm)

When access chamber cover (manhole cover) is completely removed

I don't gas test

South Western Region

Area B

Area L

Area M

Area S

Number of people who test when the cover is partly lifted (e.g. 50mm)

5

3

4

5

Number of people who test when the cover is completely removed

7

6

11

2

Number of people who replied "I don't gas test"

0

1 (+)

1 (#)

1 (*)

Number of people who did not reply to this question

0

0

0

0

Percentage compliance

42%

33%

25%

63%

(+) This person replied in Question 14 that he was not a gas tester: reply was not counted as non-compliance.

(#) This person did not reply to Question 14: counted as non-compliance.

(*) This person's reply to Question 14 indicated that he was a gas tester: counted as non-compliance.

The chi square test was not done because the number of expected replies for compliance for Areas B, L and S were 4.53, 3.40 and 3.02 respectively.
 


Reply to question 8:

If your answer to Question 7 is the first one, where do you place the gas meter on a windy day?

Upwind of access chamber cover (manhole cover)

Downwind of access chamber cover (manhole cover)

It does not matter

South Western Region

Area B

Area L

Area M

Area S

Number of people who gas test when the cover is partly lifted (e.g. 50mm) and who place the gas meter upwind on a windy day

1

0

1

1

Number of people who gas test when the cover is partly lifted (e.g. 50mm) and who place the gas meter downwind on a windy day

4

2

1

4

Number of people who gas test when the cover is partly lifted (e.g. 50mm) and who replied that it doesn't matter where the gas meter is placed

0

1

2

0

Percentage compliance

25%

20%

6%

50%

 


Reply to question 16:

Have you seen the Technical Information Sheet "Precautions when removing access covers in the possible presence of flammable gases/vapours"?

South Western Region

Area B

Area L

Area M

Area S

Number of people who have seen the Tech Info Sheet

5

3

3

4

Number of people who have not seen the Tech Info Sheet

7

7

13

4

Number of people who did not reply to this question

0

0

0

0

Percentage "reached"

42%

30%

19%

50%

 
The chi square test was not done because the number of people expected to have seen the Technical Information Sheet for Areas B, L and S were 3.9, 3.3 and 2.6 respectively.
 


8.5 Results of survey - Southern Region

This Region has one Business Unit Manager to whom two Local Waste Water Managers report. The Region is divided into three Areas referred to as Area B, Area P/K and Area W.

Southern Region

Area B

Area P/K

Area W

Number of people

10

19

8

Number replying to questionnaire

6

9

3

Percentage reply

60%

47%

38%

 


Reply to question 1:

How far do you keep ignition sources away from access chambers (manholes) when removing the covers?

1 metre

3 metre

6 metre

Southern Region

Area B

Area P/K

Area W

Number of people who keep ignition sources at least 3m away

6

9

3

Number of people who do not keep ignition sources 3m away

0

0

0

Number of people who did not reply to this question

0

0

0

Percentage compliance

100%

100%

100%

 


Reply to question 2:

Do you wet the seal of access chambers covers (manhole covers) before lifting?

always

sometimes

never

Southern Region

Area B

Area P/K

Area W

Number of people who always wet the seal

1

0

1

Number of people who sometimes wet the seal

5

3

1

Number of people who never wet the seal

0

6

1

Number of people who did not reply to this question

0

0

0

Percentage compliance

17%

0%

33%

 


Reply to question 4:

How do you lift/remove an access chamber cover (manhole cover)?

"Gatic" lifter

manhole key

"Liftrite" lifter

Shovel

Lifting chisel

Southern Region

Area B

Area P/K

Area W

Number of people who only use "Gatic" lifters and/or "Liftrite" lifters

0

0

0

Number of people who do not always use "Gatic" lifters and/or "Liftrite" lifters

6

8

3

Number of people who did not reply to this question

0

1

0

Percentage compliance

0%

0%

0%

 


Reply to question 7:

If you do a gas test, at what stage do you test?

When access chamber cover (manhole cover) is partly lifted (e.g. 50 mm)

When access chamber cover (manhole cover) is completely removed

I don't gas test

Southern Region

Area B

Area P/K

Area W

Number of people who test when the cover is partly lifted (e.g. 50mm)

6

6

2

Number of people who test when the cover is completely removed

0

3

1 (*)

Number of people who replied "I don't gas test"

0

0

0

Number of people who did not reply to this question

0

0

0

Percentage compliance

100%

67%

67%

 

(*) This person replied that he tested all access chamber covers before taking the cover off. This was counted as non- compliance because a tight cover would not allow gas to escape to be able to be detected.

 


Reply to question 8:

If your answer to Question 7 is the first one, where do you place the gas meter on a windy day?

Upwind of access chamber cover (manhole cover)

Downwind of access chamber cover (manhole cover)

It does not matter

Southern Region

Area B

Area P/K

Area W

Number of people who gas test when the cover is partly lifted (e.g. 50mm) and who place the gas meter upwind on a windy day

1

0

0

Number of people who gas test when the cover is partly lifted (e.g. 50mm) and who place the gas meter downwind on a windy day

4

6

1

Number of people who gas test when the cover is partly lifted (e.g. 50mm) and who replied that it doesn't matter where the gas meter is placed

0

0

0

Percentage compliance

67% (+)

67%

33%

(+) One person replied that he did a gas test when the cover was partly lifted (e.g. 50mm), but did not reply to Question 8: reply was counted as non-compliance.

 


Reply to question 16:

Have you seen the Technical Information Sheet "Precautions when removing access covers in the possible presence of flammable gases/vapours"?

Southern Region

Area B

Area P/K

Area W

Number of people who have seen the Tech Info Sheet

6

3

0

Number of people who have not seen the Tech Info Sheet

0

6

1

Number of people who did not reply to this question (*)

0

0

2

Percentage "reached"

100%

33%

0%

(*) Counted as non-compliance.

 


8.6 Results of survey - All Regions

All Regions

Northern

North Western

Central

South Western

Southern

Number of people

113

104

145

89

37

Number replying to questionnaire

47

48

110

46

18

Percentage reply

42%

46%

76%

52%

49%

 


Reply to question 1:

How far do you keep ignition sources away from access chambers (manholes) when removing the covers?

1 metre

3 metre

6 metre

All Regions

Northern

North Western

Central

South Western

Southern

Number of people who keep ignition sources at least 3m away

46

47

104

45

18

Number of people who do not keep ignition sources 3m away

0

1

4

1

0

Number of people who did not reply to this question (+)

1

0

2

0

0

Percentage compliance

98%

98%

95%

98%

100%

(+) Counted as non-compliance.

 


Reply to question 2:

Do you wet the seal of access chamber covers (manhole covers) before lifting?

always

sometimes

never

All Regions

 

Northern

North Western

Central

South Western

Southern

Number of people who always wet the seal

4

12

23

7

2

Number of people who sometimes wet the seal

32

31

85

26

9

Number of people who never wet the seal

11

5

0

13

7

Number of people who did not reply to this question (+)

0

0

2

0

0

Percentage compliance

9%

25%

21%

15%

11%

(+) Counted as non-compliance.

Statistical analysis

If the replies of all Regions are taken into consideration, the expected number of replies for people who wet the seal for Southern Region is 3.21. This is considered too low for a chi square test. If Southern Region is excluded, a chi square test can be done on the other four Regions because the number of expected replies is greater than 5 for each cell. The observed and expected replies are shown below.

Northern

North Western

Central

South Western

Numbers who wet the seal

4(obs)

8.61(exp)

12(obs)

8.80(exp)

23(obs)

20.2(exp)

7(obs)

8.43(exp)

Numbers who do not wet the seal

43(obs)

38.4(exp)

36(obs)

39.2(exp)

87(obs)

89.8(exp)

39(obs)

37.6(exp)

 

c2 = (O-E)2

---------

E

= (4-8.61)2 + (12-8.80)2 + (23-20.16)2 + (7-8.43)2

------------ ------------ ------------- -----------

8.61 8.80 20.16 8.43

 

= (43-38.39)2 + (36-39.20)2 + (87-89.84)2 + (39-37.57)2

------------- ------------- ------------- -------------

38.39 39.20 89.84 37.57

= 5.234

There is no significant difference at the p = 0.05 level of significance between the Regions in the number of people who wet the seal.


Reply to question 4:

How do you lift/remove an access chamber cover (manhole cover)?

"Gatic" lifter

manhole key

"Liftrite" lifter

Shovel

Lifting chisel

All Regions

Northern

North Western

Central

South Western

Southern

Number of people who only use "Gatic" lifters and/or "Liftrite" lifters

7

1

3

5

0

Number of people who do not always use "Gatic" lifters and/or "Liftrite" lifters

40

47

107

41

17

Number of people who did not reply to this question (+)

0

0

0

1

1

Percentage compliance

15%

2%

3%

11%

0%

(+) Counted as non-compliance.

The chi square test was not done because the number of people expected to only use the "Gatic" lifter and/or the "Liftrite" lifter in Northern, North Western, South Western and Southern Regions were 2.79, 2.84, 2.79 and 1.07 respectively.
 


Reply to question 7:

If you do a gas test, at what stage do you test?

When access chamber cover (manhole cover) is partly lifted (e.g. 50 mm)

When access chamber cover (manhole cover) is completely removed

I don't gas test

All Regions

Northern

North Western

Central

South Western

Southern

Number of people who test when the cover is partly lifted (e.g. 50mm)

4

36

18

17

14

Number of people who test when the cover is completely removed

39

11

71

26

4

Number of people who replied "I don't gas test"

2

0

17

3

0

Number of people who did not reply to this question (+)

2

1

4

0

0

Percentage compliance

9%(a)

25%

19%(b)

38%(c)

78%

(+) Counted as non-compliance

(a) Two people were not counted as non-complying because they were not gas testers.

(b) Thirteen people were not counted as non-complying because they were not gas testers.

(c) One person was not counted as non-complying because he was not a gas tester.

Statistical analysis

The number of expected replies in each cell was greater than 5. A chi square test was done to see if there was a significant difference between the Regions. The observed and expected replies are shown below.

Northern

North Western

Central

South Western

Southern

Numbers who test when partly lifted

4 (obs)

15.83 (exp)

36 (obs)

16.89 (exp)

18 (obs)

34.12 (exp)

17 (obs)

15.83 (exp)

14 (obs)

6.33 (exp)

Numbers who do not test when partly lifted

41 (obs)

29.17 (exp)

12 obs)

31.11 (exp)

79 (obs)

62.88 (exp)

28 (obs)

29.17 (exp)

4 (obs)

11.7 (exp)

c2 = (O-E)2

------

E

= (4-15.83)2 + (36-16.89)2 + (18-34.12)2 + (17-15.83)2 + (14-6.33)2

------------- -------------- ------------- -------------- -------------

15.83 16.89 34.12 15.83 6.33

 

= (41-29.17)2 + (12-31.11)2 + (79-62.88)2 + (28-29.17)2 + (4-11.67)2

------------- -------------- ------------- -------------- -------------

29.17 31.11 62.88 29.17 11.67

= 73.22

There is a significant difference at the p = 0.05 level of significance between the Regions in compliance with the procedure to gas test when the cover is partly lifted.
 


Reply to question 8:

If your answer to Question 7 is the first one, where do you place the gas meter on a windy day?

Upwind of access chamber cover (manhole cover)

Downwind of access chamber cover (manhole cover)

It does not matter

All Regions

Northern

North Western

Central

South Western

Southern

Number of people who gas test when the cover is partly lifted (e.g. 50mm) and who place the gas meter upwind on a windy day

0

7

2

3

1

Number of people who gas test when the cover is partly lifted (e.g. 50mm) and who place the gas meter downwind on a windy day

4

28

15

11

11

Number of people who gas test when the cover is partly lifted (e.g. 50mm) and who replied that it doesn't matter where the gas meter is placed

0

1

1

3

0

Percentage compliance

8.5%

58%

14%

24%

61%

The chi square test was not done on the five Regions because the number of expected replies for people who were non-complying for Northern, Central, South Western and Southern Regions were 0.83, 3.72, 3.52 and 2.48 respectively. Also the expected number of replies for people who complied in Northern was 3.17.
 


Reply to question 16:

Have you seen the Technical Information Sheet "Precautions when

removing access covers in the possible presence of flammable gases/vapours"?

 

All Regions

Northern

North Western

Central

South Western

Southern

Number of people who have seen the Tech Info Sheet

13

26

54

15

9

Number of people who have not seen the Tech Info Sheet

32

18

54

31

7

Number of people who did not reply to this question (+)

2

4

2

0

2

Percentage "reached"

28%

54%

49%

33%

50%

(+) Counted as non-compliance.

Statistical analysis

All expected replies were greater than 5 for each cell. The number of observed and expected replies are shown below. A chi square test was done to see if there was any significant difference between the Regions

Northern

North Western

Central

South Western

Southern

Numbers who have seen the Tech Info Sheet

13 (obs)

20.44 (exp)

26 (obs)

20.88 (exp)

54 (obs)

47.84 (exp)

15 (obs)

20 (exp)

9 (obs)

7.83 (exp)

Numbers who have not seen the Tech Info Sheet (*)

34 (obs)

26.56 (exp)

22 (obs)

27.12 (exp)

56 (obs)

62.16 (exp)

31 (obs)

26 (exp)

9 (obs)

10.17 (exp)

(*) People who did not reply to this question are included here.

c2 = (O-E)2

-------

E

= (13-20.44)2 + (26-20.88)2 + (54-47.84)2 + (15-20)2 + (9-7.83)2

-------------- -------------- -------------- ---------- -----------

20.44 20.88 47.84 20 7.83

+ (34-26.56)2 + (22-27.12)2 + (56-62.16)2 + (31-26)2 + (9-10.17)2

------------- -------------- -------------- ----------- ------------

26.56 27.12 62.16 26 10.17

= 10.94

There is a significant difference at the p = 0.05 level of significance between the Regions in the number of people who saw the Technical information Sheet.
 

9. Discussion

The results of a survey on confined spaces was used to determine compliance with the Technical Information Sheet. It is appropriate to first consider the replies to Question 16 to determine the number of people who actually received the Technical Information Sheet.

9.1 People who received the Technical Information Sheet

Of the 269 people who replied, a total of 117 received the Technical Information Sheet and 10 did not reply to this question. The 117 people represent 43.5% of the people who replied to the survey. Statistical analysis of the replies to question 16 showed that there was a significant difference (p<0.05) between Regions with regards to the distribution to sewerage maintenance people of the Technical Information Sheet. In Northern Region and South Western Region only 28% and 33% respectively received the Technical Information Sheet. While the reasons for the low percentages were not known, it is possible that a change of management in Northern at Business Unit Manager level may have created discontinuity in the flow of information. A higher percentage of people in North Western, Central and Southern Regions receiving the Technical Information Sheet may have been due to these three Regions being located in more highly industrialised areas where the frequency of illegal trade waste discharges (including flammable material) is higher.
There was also a significant difference (p<0.05) in the distribution of the Technical Information Sheet within Central Region. This Region has three Areas. A greater percentage of people in the more highly industrialised areas, that is, Area E and Area C/E, received the Technical Information Sheet, 61% and 55% respectively. Only 31% of people in the other area, Area St.G, received the information. The demographic data for these three Areas showed that education level was no obstacle to people seeing the information. Although Area E had a lower level of education, it had the best response.
The results for question 16 were disappointing considering the effort that had been made to make people aware of fire and explosion hazards in sewers. The draft Technical Information Sheet had been circulated to Business Unit Managers for comment and input. When finalised, they were given copies for distribution among their respective Areas. Distribution also occurred in other areas such as the Sewage Treatment Plants and the Incident Management Unit of the Provider. Subsequent to distribution of the Technical Information Sheet, the Incident Management Unit conducted training sessions at Local Waste Water Management level throughout the five Regions on procedures to be taken in the event that natural gas was found in the sewers. Although these training sessions concentrated on natural gas (essentially methane), the procedures involved in the Technical Information Sheet were reinforced at each of these sessions.
 

9.2 Compliance with the procedure "Where practicable, keep potential sources of ignition at least 3 metres away from the chamber e.g. 2-way radios, "motorolas", cellular phones, radios, pagers, torches, other electrical equipment, vehicles and other motorised equipment".

Question 1 assessed compliance with this procedure. Of the 269 people who replied, 260 (96.6%) complied with this procedure. This was most likely because for a number of years there had been a "Standing Order and Board Instruction" for naked lights to be kept 6 metres away from access chamber openings. This instruction stemmed back to the days before metrification when the distance was 20 feet. As an interesting aside, it is believed that the distance of 20 feet was used because it was thought that this was the distance that someone could flick a cigarette butt.
Three people did not reply to this question. Two people from Central Region replied that they did not keep ignition sources 3m away from access chambers when removing the cover. Whether these people misunderstood the question of simply ignored safety procedures is not known.
 

9.3 Compliance with the procedure "Wet the seal with water".

Question 2 assessed compliance with this procedure. Of the 269 people who replied, the number of people who strictly carried out this procedure was low, 48 (17.8%). In some cases the lack of compliance to this procedure may have been due to not having seen the Technical Information Sheet. Although for all Regions, the numbers who had seen the Technical Information Sheet was greater than the numbers complying with the procedure. The numbers complying for Northern, North Western, Central, South Western and Southern Regions were 9%, 25%, 21%, 15% and 11% whereas the numbers replying to having seen the Technical Information Sheet were 28%, 54%, 49%, 33% and 50% respectively.
A large number of people, 183 (68%) replied that they sometimes wet the seal. 30 did so "because flammable gas might be present". 19 of these people had seen the Technical Information Sheet and 11 had not. In this case the Technical Information Sheet appeared to have influenced those people who saw it.
Of the 48 who always wet the seal, 25 had seen the Technical Information Sheet and 23 had not seen it. From these observations it can be inferred that the Technical Information Sheet did not influence these people in complying with this procedure. Of the 25 people who always wet the seal and who had seen the Technical information Sheet, two did not give reasons for wetting the seal. Of the 23 who gave reasons, 11 did so to "break the seal and make lifting easier", 7 did so "because flammable gas might be present", and 5 gave both reasons. Of the 23 people who had not seen the Technical Information Sheet, 4 did not give reasons for wetting the seal. Of those who did, 6 did so "to break the seal and make lifting easier", 6 did so "because flammable gas might be present", and 7 gave both reasons. From the above observations it can be seen that the predominant reason for wetting the seal was to make removal of the access chamber cover easier.
36 people replied that they never wet the seal. 12 of these people had seen the Technical Information Sheet, 3 were in Northern and 6 in South Western. Their non-compliance with this procedure may have been due to the fact that they worked in areas which were not highly industrialised and consequently they did not expect to find flammable gases/vapours. 24 people who never wet the seal had not seen the Technical Information Sheet.
Statistical analysis of replies to Question 2, excluding Southern Region, showed that there was no significant difference between the four other Regions with respect to compliance with this procedure.
Of the 269 people who responded, only one person objected vehemently to having to wet the seal. This person worked in an area where a bucket of water would have to be carried through bushland to reach the access chamber.
 

9.4 Compliance with the procedure "To lift access cover use a lifting device which maximises the distance of the person from the cover".

Question 4 assessed compliance with this procedure. Of the 269 people who replied, 16 (5.9%) strictly complied with this procedure. Of these, 5 had seen the Technical information Sheet and the other 11 had not.
However, 180 people (66.9%) used a long handled lifting device (i.e. "Gatic" lifter or "Liftrite" lifter) at times. Several reasons could account for people not always complying with this procedure. Firstly, some may not have seen the Technical Information Sheet and therefore were not aware of the requirement to do so. Secondly, not all access chamber covers can be lifted by "Gatic" or "Liftrite" lifters. Although attachments are available so that these commercial lifting devices can be used on various access chamber covers, the availability of attachments throughout the Provider's area of operation may have been limited. Thirdly, there may have been some reluctance to use these long handled devices because the job took longer to do. People may prefer using a "manhole key", albeit with a greater risk of manual handling injury.
 

9.5 Compliance with the procedure which required testing of the atmosphere when the cover was partly lifted (50mm).

Question 7 assessed compliance with this procedure. Of the 269 people who replied, 89 (33.1%) complied with this procedure. Of this number 46 had seen the Technical Information Sheet and 39 had not. 4 people did not indicate whether they had seen the information. From these observations it can be inferred that the Technical Information Sheet did not influence people to comply with this procedure.
There was a significant difference between Regions with compliance with this procedure. Compliance for Northern, North Western, Central, South Western and Southern Regions was 9%, 25%, 19%, 38% and 78% respectively.
Question 14, although not intended to be used for the survey was found to be beneficial in elaborating replies to Question 7 as previously mentioned.
 

9.6 Compliance with the procedure "Switch on gas detector and place downwind and as close as possible to the access cover".

Question 8 assessed compliance with this procedure. Of the 269 replies, 69 people (25.7%) complied with this procedure. There was insufficient data to see if there was any significant difference between the Regions. Of the 69 people who complied, 37 had seen the Technical Information Sheet and 30 had not. Two people did not indicate whether they had seen it or not. From these observations it can be inferred that the Technical Information Sheet did not greatly influence compliance with this procedure.
 

9.7 Review of the Technical Information Sheet

From the information obtained in this research, few changes in the Technical Information Sheet are required. However, the need to warn people not to drop the cover, if flammable gases/vapours are present, is required.

Procedure 1

"Where practicable, keep potential sources of ignition at least 3 metres away from the chamber e.g. 2-way radios, "motorolas", cellular phones, radios, pagers, torches, other electrical equipment, vehicles and other motorised equipment"
No changes are required to this procedure. In the literature, only Foster (1979) specifies a safe distance for ignition sources to be kept away from access chambers when the cover is removed. He specifies a distance of at less 3 metres which is the same as specified in the Technical Information Sheet. The survey showed that most people (96.6%) complied with this requirement and no one indicated having difficulties meeting the requirement.

Procedure 2

"Wet the seal with water"
From the research reviewed, the need to wet the seal is not necessary. However, the information gained in the survey makes retention of Procedure 2 a necessity. Firstly, a number of people commented that the use of a hammer and chisel, to loosen covers which are wedged into place, generates sparks. Mention has already been made that sparks generated by hand tools have sufficient energy to ignite flammable gases/vapours-air mixtures. Wetting the seal would improve the barrier between potentially explosive gases/vapours-air mixtures in the sewer and the surface where sparks are being generated. Secondly, in reply to Question 4, three people mentioned that they used a winch attached to the back of a truck to lift covers. With the winch system there is a possibility, with a stuck cover, for it to suddenly "pop out" when the tension in the cable exceeds the frictional forces holding the cover in place. In this situation, the velocity of the cover relative to its surround may be sufficiently high to generate local heating/or sparks sufficient to cause ignition.
The question arises whether it should always be mandatory to wet the seal or only be mandatory when a hammer and chisel and/or a winch system is used. The replies to the questionnaire indicated that the requirement to wet the seal was not too onerous because 85.8% of people replied that they always (17.8%) or sometimes (68%) wet the seal. Only one person objected to having to wet the seal because water had to be carried through bushland to the access chamber. Since water is already carried on trucks for people to wash their hands after a job, it does not seem unreasonable for water to be also carried to wet seals. It is considered appropriate for this procedure to be retained.

Procedure 3

"Switch on gas meter and place downwind and as close as possible to the access cover"
No changes are required to this procedure.

Procedure 4

"To lift access cover use a lifting device which maximises the distance of the person from the cover"
This procedure should be retained purely on the grounds that it minimises the risk of manual handling injuries. Although only 16 people (5.9%) used long handled lifting devices all the time, 66.9% used these devices at times.

Procedure 5

"Slowly lift cover about 50 mm"
No changes are required.

Alter Step 6 in the Technical Information Sheet

As a result of the literature review, the third sentence in Procedure 6 (Appendix A) should now read:
"If concentration is 40% LEL or greater, replace cover gently; do not replace cover if directed by the site manager."

 

10. Conclusion

The literature review has shown that fire and explosions in sewers and associated structures occur on a regular basis in the U.S.A. Since a similar incident has already occurred in Australia and there is the potential for others to occur, it is essential that Providers of sewerage facilities have established procedures in place to avoid ignition of flammable gases/vapours-air mixtures when accessing sewers.
The literature search has confirmed that the procedures outlined in the Technical Information Sheet are appropriate steps to take when removing access chamber covers in the possible presence of flammable gases/vapours. However, the literature search has also highlighted the need to include a step warning people not to drop the cover if the concentrations of flammable gases/vapours-air mixtures are near or above the lower explosive limit.
Total Quality Management is currently being actively pursued by many organisations. Without a quality assurance program, an organisation has no idea on how well safety programs are being implemented and managed. The survey showed that by simply distributing the Technical Information Sheet only 43.5% of people received it. Training of every sewerage maintenance person on the safety procedures would be required for better compliance. In addition, training is required to rectify some of the faulty procedures found from the survey, such as; people not keeping ignition sources 3m away from access chambers when removing the covers and people placing gas meters upwind when taking readings.
Compliance with the procedures in the Technical Information Sheet varied greatly, from 5.9% for people who always used a long handled lifting device to 96.6% for people who kept ignition sources at least 3 metres away from access chambers when the cover was being removed. However, if the number of people who sometimes carry out the requirements of the Technical Information Sheet is considered, the potential compliance figures look much better. For example, the number of people who sometimes used a long handled lifting device was 66.9%, and the number of people who sometimes wet the seal was 68%.
The results show good compliance with parts of the Technical Information Sheet. Further work, such as training, is required to ensure that sewerage maintenance people comply with all aspects of the Technical Information Sheet.

 

11. References

Abbott, J. 1993, Enclosed Wastewater Treatment Plants - Health and Safety Considerations, Research Report FR/W 0001, Foundation for Water Research, Allen House, The Listons, Liston Road, Marlow, Buckinghamshire, Nov, pp. 3-17 and 6-2.

Barker, R. 1991, "Plotting Gunpowder Safety in a Fireworks Factory", The Safety & Health Practitioner, November, pp. 22-24.

Billinge, K. 1979, Fire Prevention Science and Technology, vol.24, pp.13-19 vide Powell, F. 1986, "Can Non-Sparking Tools and Materials Prevent Gas Explosions," Gas-Wasser-Abwasser, 66. Jahrgang no.6, pp. 419-428.

Bishop, P. 1992, "Undetected leak triggers Mexico sewer explosion", Underground, no.23, pp.4-5.

Bowden, F.P. & Tabor, D. 1954, The Friction and Lubrication of Solids, 2nd Ed, Clarendon Press, Oxford vide Powell, F. 1969, "Ignition of Gases and Vapors", Industrial and Engineering Chemistry, vol 61, no. 12, pp. 29-37.

Bowker, R.P.G; Smith, J.M. & Webster, N.A. 1989, Odor and Corrosion Control in Sanitary Sewerage Systems and Treatment Plants, Hemisphere Publishing Corporation, New York, London, pp. 14.

Department of Scientific and Industrial Research and Fire Offices' Committee Joint Fire Research Organisation on Industrial Fires and Explosions. 1963, "Report on 'Non-sparking' Hand Tools", Journal of the Institute of Petroleum, vol. 49, no. 474.

Drager 1994, Drager-Tube Handbook, 9th edn, Dragerwerk Aktiengesellschaft Lubeck, Germany, pp. 146-147.

Feldman, H.V. 1977, "Gas Hazards And Their Detection For Safety Purposes", Prog. Wat. Tech., vol.9, nos. 5/6, pp. 137-143.

Fingas, M.F; Hughes, K.A. & Bobra, A.M. 1989, The Behaviour of Dispersed and Nondispersed Fuels in a Sewer system, Oil Dispersants: New Ecological Approaches, ASTM STP 1018, Flaherty, L.M. (ed), American Society for Testing and Materials, Philadelphia, pp. 274-289.

Foster, W.S. 1979, "How to keep workers safe", The American City & Country, April, pp. 32.

Gerardi, M.H. 1982, "A Review of Dangerous Gases in Sanitary Sewers", Public Works, Oct, pp. 34-36.

Guo, Q. & Song, C.C.S. 1991, "Dropshaft Hydrodynamics Under Transient Conditions", Journal of Hydraulic Engineering, vol. 177, no.8, pp. 1042-1055.

Hurlbut, C.S. 1952, Dana's Manual of Mineralogy, John Wiley & Sons, Inc. London, Chapman & Hall Ltd, pp.111.

Instrumentation In The Water Industry. 1976, "Use of intrinsically safe instrumentation", Water Services, June, pp. 356-357.

Jones, A.T. 1968, "The control of dust in tunnelling and excavation work" in Proceedings of the First Australian Pneumoconiosis Conference, Joint Coal Board, Sydney, pp. 357 & pp. 363.

Kletz, T. 1993, Lessons From Disasters, How Organisations Have No Memory and Accidents Recur, Institution of Chemical Engineers, Warwickshire, U.K, pp.124.

Kocherga, N.G. 1969, "The mechanism of frictional sparking of steel and rocks", Trudy Gosud. Makeevskii n-i Inst. po Bezopasnosti Rabot v gornoi prom, vol. 19, pp.185-194 (English translation).

Larsen, R.J. 1975, Statistics for the Health Sciences, Charles E. Merrill Publishing Company, Columbus, Ohio, pp. 245-247.

Lewis, W.T; Smith, P.R. & Powell, F. 1983, "Paper D1", 20th International Conference on Mine Safety Research, October vide Powell, F. 1986, "Can Non-Sparking Tools and Materials Prevent Gas Explosions", Gas-Wasser-Abwasser, 66. Jahrgang no.6. (English translation).

National Institute of Occupational Safety and Health (NIOSH). 1985, Request for assistance in preventing hazards in the use of water spray (fog) streams to prevent or control ignition of flammable atmospheres, Robert A. Taft Laboratories, 4676 Columbia Parkway, Cincinnati, OH 45226, USA, July.

National Fire Protection Agency (NFPA). 1982, "Analysis of Cause of Explosions", Recommended Practice for the Control of Flammable and Combustible Liquids and Gases in Manholes, Sewers, and Similar Underground Structures, Appendix B, National Fire Protection Association, USA, pp. 328-9 to 328-11.

National Fire Protection Agency (NFPA) 820, 1992, Recommended Practice for Fire Protection in Wastewater Treatment and Collection Facilities, National Fire Protection Association, USA, pp.16.

New Technology. 1994, "There's No Beating CCTV In The Pipeline", Australian Plumbing Industry, August/September, pp. 18-19.

Palmer, K.N. 1960. Engineering, July 15 vide Powell, F. 1986, "Can Non-Sparking Tools and Materials Prevent Gas Explosions," Gas-Wasser-Abwasser, 66. Jahrgang no.6, pp. 419-428.

Perry, R.H. 1984, Perry's Chemical Engineers' Handbook, 6th ed, McGraw-Hill Book Company, pp. 3-72.

Pilkington, N.H. & Bridger, J.S. 1994, "Flow-Proportional Sampling of Sewage", Chemistry in Australia, August, pp. 438-439.

Polit, B. & Hungler, D. 1987, Nursing Research: Principles and Methods, Philadelphia, USA vide Herron, E. 1994, "OH: perceptions and awareness by NHS CHC employees", Occupational Health, Dec, pp.408-411.

Powell, F. & Billinge, K. 1975, "The Frictional Ignition Hazard associated with Colliery Rocks", The Mining Engineer, July, pp. 527-533.

Powell, F. 1969, "Ignition of Gases and Vapors", Industrial and Engineering Chemistry, vol 61, no. 12, pp. 29-37.

Powell, F. 1984, "Ignition of Gases and Vapours by Hot Surfaces and Particles - A Review", 9th International Symposium on the Prevention of Occupational Accidents and Diseases in the Chemical Industry, Luzern, 5-7. Juni, pp. 267-292.

Powell, F. 1986, "Can Non-Sparking Tools and Materials Prevent Gas Explosions," Gas-Wasser-Abwasser, 66. Jahrgang no.6, pp. 419-428. (Article in English)

Powell, F; Billinge, K. & Cutler, D.P. 1975, The ignition of methane-air by machine picks cutting into rock, Safety in Mines Research Establishment Health and Safety Executive, Buxton Great Briton.

Powell, F. & Quince, B.W. 1972, "The kinetics of impact in drop-weight experiments", J. Phys.D: Appl. Phys, vol.5, pp. 1540-1553.

Prentiss, D. & Laughlin, J. 1995, "NFPA 820 Being Revised, Upgraded", Water World, vol 11, no.1, pp.1 & 8.

Rae, D. et al. 1964, Safety Mines Res. Rept, no. 224, Safety in Mines Research Establishment, Sheffield vide Powell, F. 1969, "Ignition of Gases and Vapors", Industrial and Engineering Chemistry, vol.61, no.12, pp. 29-37.

Rae, D. 1964, The Role of Quartz in the Ignition of Methane by the Friction of Rocks, Research Report No. 223, Safety in Mines Research Establishment, Sheffield, UK, pp.23.

Reuters. 1992, File 772, Textline, United Kingdom News 1988-1992/Jul 02 (C).

Sawyer, C.H. & McCarty, P.L. 1978, Chemistry For Environmental Engineering, McGraw-Hill Book Company, New York, pp. 494-495.

Sedlak, R. 1991, Phosphorus and Nitrogen Removal from Municipal Wastewater - Principles and Practice, Lewis Publishers, US, pp. 37.

Silver, R.S. 1937, Phil.Mag, vol.23, no.7, pp. 633-57 vide Powell, F. 1969, "Ignition of Gases and Vapors", Industrial and Engineering Chemistry, vol.61, no.12, pp. 29-37.

Standards Australia HB13/NEEITC 181-1-1992. 1992, Electrical Equipment for Hazardous Areas.

Standards Australia 2380.1. 1989, Electrical equipment for explosive atmospheres - Explosion-protection techniques, Part 1: General requirements, pp.7.

Standards Australia 1940. 1993, The storage and handling of flammable and combustible liquids, pp. 68.

Strange, J. & MacKenzie-Wood, P. 1985, A Manual On Mines Rescue And Gas Detection, Weston & Co. Publishers Pty. Ltd. Kiama, Australia, pp. 120-121.

Sydney Water Board. 1986, "Gas Theory In S.P.S. Blast", Aquarian, October, pp. 3.

Tweeddale, H.M. 1994, Review of Reliability and Risk Management at Sydney Water Board Ocean Outfall Sewerage Treatment Plants, Australian Centre of Advanced Risk and Reliability Engineering Ltd.

Voigtsberger, P. 1955, Bundesarbeitblatt, pp. 191-198, HSE (SMRE) Transl. 5230 vide Powell, F. 1986, "Can Non-Sparking Tools and Materials Prevent Gas Explosions," Gas-Wasser-Abwasser, 66. Jahrgang no.6, pp. 419-428.

Ward, C.R; Cohen, D; Panich, D; Crouch, A; Schaller, S. & Dutta, P. 1990, Assessment Of Gas Ignitability Potential By Frictional Effects From Australian Coal Mine Rocks, Commonwealth of Australia National Energy Research, Development and Demonstration Program, End of Grant Report Number 1010, September.

Wolfs, J.R. 1959, "Explosion Prevention", Sewage and Industrial Wastes, vol 31, no. 3, pp. 321-323.


Photograph 1
An Access Chamber (Manhole) Cover Key Being Used To Lift A Cover
Photograph 2
A Modified Access Chamber Cover Key In Use
(Provided by: Margaret Ratcliffe)
Photograph 3
Use Of An Access Chamber Cover Key By Two People
(Provided by: Morris Nakhla)
Photograph 4
A Hammer And Chisel Being Used To Loosen An Access Chamber Cover
(Provided by: Morris Nakhla)
Photograph 5
A "Gatic" Lifter Being Used To Lift A Square Access Chamber Cover
(Provided by: Morris Nakhla)
Photograph 6
A "Gatic" Lifter Being Used To Lift A Round Access Chamber Cover
(Provided by: Morris Nakhla)
Photograph 7
A "Liftrite" Lifter In Use
(Provided by: Manufacturer)