Control of Pollution (Oil Storage) (England) Regulations 2001 Every year more than 5000 oil incidents are reported to the Environment Agency. In 2000 one-sixth of all pollution incidents affecting the evnironment involved oil. Most incidents were caused by oil leaking from tanks either during storage or delivery. The new Oil Storage Regulations will help us to stop these incidents by requiring tank owners to provide a secondary containment facility, such as a bund or drip tray to prevent oil escaping into the water environment. Anyone storing more then 200 litres of oil above ground at an industrial, commercial or institutional site, or more than 3500 litres at a domestic site will be affected by these regulations. They cover factories, shops, offices, hotels, schools, public sector buildings and hospitals. The regulations apply only in England, but Wales and Scotland are likely to follow suite.
Which types of oil are covered? All types of oil, with the exception of waste oil, are covered by these regulations including petrol, diesel, vegetable, synthetic and mineral oil. Wast oil is covered by the Waste Management Licensing Regulations. It is important to note that in the case of flammable liquids, such as petrol, additional health and safety requirements may also apply.
What are the standards? Tanks, drums or other containers must be strong enough to hold the oil without leaking or bursting.
If possible, the oil container must be positioned away from any vehicle traffic to avoid damage from collision. A bund or drip tray must be provided to catch any oil leaking from the container or its ancillary pipework and equipment. Where more than one container is stored, the bund should be capable of storing 110% of the largest tank or 25% of the total storage capacity, whichever is the greater. The bund base and wlls must be impermeable to water and oil and checked regularly for leaks. Any valve, filter, sight gauge, vent pipe or other ancillary equipment must be kept within the bund when not in use.
No drainage valve may be fitted to the bund for the purpose of draining out rainwater. Above-ground supported.
pipework
should
be
properly
Underground pipework should be protected from physical damage and have adequate leakage detection. If mechanical joints must be used, they should be readily accessible for inspection.
A number of other detailed requirements are included in the regulations, such as the positioning of sight guages, fill points, vent pipes and other ancillary equipment.
When do the regulations come into effect? Now oil stores will have to comply from 1st Masrch 2002. Existing oil stores “at significant risk” will have to comply within two years, that is from 1st September 2003.
All remaining existing oil stores must comply wiithing four years, that is from 1st September 2005. In general, an oil store will be considered to be at “significant risk” if it is located within 10 metres of a watercourse or 50 metres of a well or borehole.
Are mobile containers covered? Yes. The regulations apply to drums greater than 200 litres and to mobile bowsers. Many self-bundowsers are now available. Those that are not bunded will need to be kept in a bunded area or a drip tray when in use. For single drums, a drip tray with a capacity of 25% is acceptable.
How are the regulations enforced? The Environment Agency is responsible for enforcing these regulations throughout England. Should your oil storage facilities be inadequate, the Agency will provide advice and guidance to assist you with compliance. However, if you fail to act, the Agency may serve a notice requiring that the facilities be brought up to standard. Failure to comply with a notice is a criminal offience any may result in prosecution.
The regulations do not apply: at premises used wholly or mainly as a single private dwelling storing less than 3500 Itres. at premises used for refining oil or its onward distribution; any oil stored in a building or wholly underground; to farms - the storage of agricultural fuel oil on farms is subject to the 1991 Silage, Slurry and Agricultural Fuel Oil Regulations; to waste oil.
Do you need further advice? The Environment Agency publishes a Pollution Prevention Guidance note on oil storage, PPG2 and PPG26, that provides practical advice, which, if followed, will ensure compliance with the regulations.
Leak Detection and Leakage Prevention The Dependable System by Jost Berg.
Leak detection and leakage prevention There are many methods used in the oil industry to detect leaks on tanks or pipes but only a few of those available some are able to prevent leakage from contaminating the environment. Basically all methods, which use the stored product somehow itself for leak detection purposes are more or less are not able to prevent leakage into the environment. Examples are: Tank gauging systems doing static or dynamic reconciliation will detect a loss of product, which is on single-walled tanks and cause environmental pollution at the same time. Necessarily working with measuring tolerances in the system and trying to determine which losses are caused by vapour or by leaks, they have to allow for specified amounts of liquid disappearing without recognition. These small amounts can accumulate through the years to cause pollution underground and nowadays these are becoming less acceptable especially to real estate owners. Monitoring wells will detect a leak and a leakage as the hydrocarbon sensor placed beside the tank gets into contact with the stored product. Environmental pollution and possibly high decontamination cost are clearly indicated by using this system. Sensors in the interstitial space or in the interstitial space access pipe are located there to wait for product or ground water. They will 'trigger the alarm in case of contact with liquid. In case of a leak in the inner wall they will probably recognize the product but they will not give any indication about the integrity of the outer wall. It remains still a risk of product leaking into the environment, especially reflecting that most tanks get a leak in the outer wall first. Line leak detection systems look for unacceptable pressure drops in the line during a specified measuring time. In case there are those pressure drops, a leakage and a pollution is doubtless detected. The only safe method to detect leaks and to prevent leakage and environmental pollution is firstly to use double walled tanks and pipes and secondly to monitor permanent and 100 percent the integrity of the interstitial space.
These systems have been in use for some years now and there has been excellent experience in keeping service stations leak free. In 1956 German law claimed that groundwater has to be protected against environmental pollution. Leak detection systems for single-walled tanks have been developed and used up to the 1980's. Finding out that these systems still have some weak points an upgrade in law in the 70's required double-walled tanks and leak detection systems which will give an alarm before any product can enter the environment. Technical solutions were soon produced by the industry and technical rules to define the function and product performances were developed by the authorities. Within a short time a new standard was realized, requiring and defining pressure, vacuum and liquid leak detection systems. These standards have been taken over by other countries like Switzerland and Austria quite parallel to the development in Germany. Within the last 15 years requirements for double walled tanks and pipes equipped with vacuum, pressure or liquid leak detection system spread throughout Europe. The European standard on leak detection soon to be published recognizes the superiority of the vacuum and pressure systems by making these systems class one where no product will enter the environment before a leak is detected. But also the technical differences between liquid and pressure systems have been realized during this period. Liquid leak detection systems work by filling the interstitial space with leak detection liquid and locating a header tank with a sensor inside on top of the tank. In case of a leak, leak detection liquid will enter the inner tank or the environment, causing a liquid level drop in the header tank which is recognized by a sensor.
It is necessary to place the header tank higher than the ground water table, which might cause additional installation cost in areas with high ground water tables. In cases of stored liquids with higher densities the header tank has to be installed in a corresponding height. Limits are given by the pressures the interstitial space resists to.
Part of the functioning of the liquid system is, that the stored product and/or the environment will be contaminated with the leak detection liquid in case of a leak. In many European countries the liquid systems are not allowed anymore to be installed new as the used liquid itself is defined to be a water endangering product. (e.g. Ethyleneglycol with fungicides) Technical problems occured, are that the liquid could gel by topping up with the wrong liquid or that zinc plated steel used for the piping caused crystalisation of the liquid. Both problems cause a tank related failure of the leak detection system and a necessary replacement of the tank.
Pressure or Vacuum leak detection systems based on air or inert gas will not influence the tank and the environment in any negative way. They produce a pressure difference to the atmospheric pressure in the interstitial space in such a way, that a leak anywhere in the interstitial space and the system itself will be definitely detected by pressure changes. These pressure changes to the alarm pressure of the system will occur before any stored product can escape into the environment. Minor atmospheric differences of the whole leak detection system consisting out of a tank, leak detector and connection lines are compensated within very tight limits by integrated pressure pumps. Vacuum leak detection sytems (also called underpressure systems) realize the following monitoring principle: The pump in the leak detector creates a operational vacuumunderpressure in the interstitial space. The operational and the vacuum alarm are adjusted higher than the pressure of the stored product to the bottom of the tank. By doing this the vacuum is high enough to suck product or groundwater in the interstitial space and up into the liquid stop valve, which is installed on top of the tank in the suction line of the leak detector. In case of a leak, air, groundwater or product is sucked into the interstitial space by the vacuum. If the volume flow of air entering the interstitial space is higher than the limited volume flow of the vacuum pump, the pressure will rise to the alarm pressure. If groundwater or stored product is sucked through a leak, the interstitial space is filled up until the liquid enters the liquid stop valve. The liquid stop valve closes and no further vacuum can be produced. Some more liquid will be sucked into the interstitial space or the measuring line, causing a rise in pressure to the alarm pressure. The optical and acoustical alarm is then sounded
Pressure leak detection systems realize the following monitoring principle: The pump in the leak detector creates a fixed operational overpressure in the interstitial space. The alarm pressure and the operational pressure are higher than the pressure (weight) of the stored product or the groundwater against the interstitial space at it‘s deepest point. In case of a leak, the compressed air will escape through the leak hole. This prevents product or groundwater entering the interstitial space.
If the volume flow of air escaping from the interstitial space is higher than the limited volume flow of the pressure pump, the pressure will drop to the alarm pressure. An optical and acoustical alarm will be released. The compressed air in the interstitial space is dried by the dry filter mounted to the leak detector. Therefore condensation of water in the interstitial space is prevented.
An overpressure valve avoids the occurrence of inadmissible overpressure in the interstitial space. Some systems use inert gas like nitrogen instead of air in case that the stored product should not get into contact with oxygen or hydrogen or to increase explosion protection measurements. Pressure and Vacuum leak detection systems for double walled tanks or pipes based on air or inert gas have been
installed and used in Europe since the middle of the 1970’s. Until today there are more than 700,000 systems in use, monitoring permanent 100% the integrity of the interstitial space of tanks and pipes. Per year a small number (estimated 1%) tanks or pipes start leaking. Approximately 75% of these have the leak in the outer wall, only 25% in the inner wall. These leaks are detected as soon as one of the two walls shows a leak. There is enough time once the alarm has been given to take action without risk of contamination. PermanantlyTtesting the integrity of the intersitial space reduces the risk of product escaping into the environment down to zero. Double-walled tanks and pipes permanent monitored with overpressure or vacuum leak detection systems are the European solution to avoid environmental pollution or fire risks from leaking product.
Jost Berg SGB Sicherungsgerätebau G m b H Hofstr. 10 D-57076 Siegen Germany www.sgb.de sgb@sgb.de
Investigation into the role of electrostatics in a vapour ignition incident during refuelling at a Snax service station. Graham Hearn, Wolfson Electrostatics, University of Southampton
During the delivery of unleaded petrol to a Snax service station on the Southend Arterial Road a small fire occurred in the manhole chamber containing the underground offset fill pipes. The manhole involved has four fill pipes and is constructed from GRP with a metal frame at the top. The pipes are plastic but have a metal transition fitting for connection to the tanker hose. When not in use the fittings have an aluminium cap with an oring seal in place. It is understood that the delivery driver had two hoses connected to the offset fill points and was delivering the product. One delivery had finished and he opened up the cap of the next fill point to commence delivery to that tank. He then went to place the cap on the pipe of the tank to which he had completed the first delivery when the ignition occurred. The fire was quickly extinguished and no one was hurt.
termination fittings and ground. In order to do this it was necessary to remove earthing cables that had been applied after the incident. In addition the electrical capacitance (C) of the fittings were also measured. From these two measurements it is possible to calculate the relaxation time for electrostatic charge on the fitting to decay. Table 1 gives the results for the fittings 2, 4, 5 and the vapour recovery point. It is interesting to note that with the exception of the diesel filling point 5 the charge relaxation times are extremely low. This includes fitting 2 where the incident occurred.
Initially it was suspected that the probable cause of the ignition was static electricity igniting an uncommonly high concentration of petrol vapour. This is not an unreasonable assumption since the metal termination on the plastic pipe was ungrounded. In order for an ignition to occur at this point however there must be an electrostatic discharge between the cap and the termination fitting of sufficient energy to exceed the minimum ignition energy of the vapour mixture. Normally the caps made from aluminium are connected to the termination fitting by means of a short length of chain but in this case the chain was broken and consequently there was no electrical continuity between the fitting and the cap. It follows that electrostatic ignition would be reliant o n either the fitting to be raised to high potential or the cap/delivery driver being raised to high potential.
Investigation
I visited the scene of the incident on Friday 3 August 2001 in order to make observations and measurements. A road tanker delivery was arranged for that day so that measurements could be undertaken during a fuel delivery. Plates 1 and 2 give an indication of the layout. The incident occurred with fill pipe/tank 2. This can be seen clearly in Plate 2 with the aluminium cap in place. The initial test performed on the termination fittings was to establish the degree of electrical isolation by measuring the resistance (R) between each of the
Three tank filling operations were monitored. Delivery of unleaded to no.2, delivery of unleaded to no.4 and delivery of diesel to no.5. Again the earthing straps which were applied after the incident were removed for the purposes of these tests and procedures applied during the incident were copied as closely as possible. All monitored deliveries were undertaken at a flow rate of approximately 1000 litres per minute through a 4 inch hose. The electrical continuity of the hoses (from end to end) were checked and found to be sound. During filling which lasted several minutes, electrostatic potentials on both the fittings and the short length of accessible plastic pipe within the manhole were monitored. Potentials on the fittings and pipe were also monitored on completion of fill and disconnection of the hose.
Fittings 2 and 4 did not rise above zero potential at any time during fill or removal of the hose. During the delivery of diesel to tank 5 a small but insignificant level of static potential (less than +2OO V) was observed on the pipe. O n completion of fill and disconnection of the hose a small reading rising to a maximum of +100 Volts was observed on the fitting. These results were not surprising in light of the rapid charge relaxation times previously calculated. Measurement of electrostatic potential on the fittings is shown in Plate 3.
Electrostatic charge generation
It is understood that the conductivity of the fuel at the time of the incident was around 400 pS/m. Based on previous work with plastic pipelines, a flow rate of around 2 m/s (equivalent to approximately 1000 L/min) would produce potentials of below 1OOOV. This would not be expected to be hazardous.
Charge accumulation The final measurements undertaken were the electrical resistivity of the ground surface around the tanker filling point and the electrical resistance between the chassis/body of the tanker and ground (through the wheels and tyres). The ground resistance was measured between two 2.5 kg electrodes placed 0.5 m apart on the ground. This yielded a resistance value of 23 M (2.3 ' 107 ) . The resistance between the body of the tanker and ground gave a value of 8 M (8.0 ’ 1 0 6 ). Both of these values can be considered significantly conductive from the point of view of electrostatic ignition hazards. Although measurements were not undertaken on the footwear and clothing worn at the time of the incident, the tanker driver present during these tests was wearing standard issue footwear from Texaco. This was measured and found to be sufficiently conductive and within the limits laid down in BS5958. Weather conditions at the time of the above measurements were good. It was warm and sunny throughout the day with a temperature of 25C and relative humidity of 56%. There was heavy rainfall throughout the previous day however the conditions within the manhole were good with only a little water present in the base of the chamber.
Analysis Based on the description of the incident a requirement for an electrostatic ignition source is that either the pipe termination fitting or the cap (or driver holding the cap) should be at high potential. This potential should be such that the resultant electrostatic discharge between the cap and the fitting exceeds the minimum ignition energy of the fuel vapour. The usual analytical approach can be applied.
This is the second requirement for electrostatic ignition. Electrostatic charge can accumulate on plastic surfaces and on isolated metal components -the latter being more dangerous since sparks can result. When the pipe system was installed it is likely that the termination fittings were electrically isolated from ground. However, the build-up of contamination on and around the fittings has now produced a dissipation path to ground -clearly indicated in Table 1. From the result in Table 1 it is apparent that electrostatic charge could not be stored on the fittings with the possible exception of diesel line 5 which exhibited far less contamination and as a consequence a higher resistance to ground. Note: It is also conceivable that due to the earlier fire a carbon deposit was generated on the external pipe surfaces thus contributing to the charge dissipation. No evidence of such a deposit was observed during this investigation and it is considered that this effect would be small compared to that of the general contamination.
Can a sufficiently energetic electrostatic discharge occur? Based on the above data the answer to this question must be no. Firstly there is insufficient charge generate in the first place due to the high conductivity of the fuel (fuel conductivity was not measured as part of this investigation but was reported by Texaco). Secondly it is difficult to see how charge could accumulate on the fitting which had a measured charge relaxation time of 0.2 milliseconds. Even if the fitting had been perfectly clean and therefore electrically isolated and based on a capacitance of 44pF an electrostatic potential of between 3.4 and 6.0kV would be required for ignition. This level of potential could not be envisaged with fuel of this conductivity at these flow rates.
Presence of flammable atmosphere The flammable range of petroleum spirit vapour in air is 1% - 7% by volume. The minimum ignition energy varies greatly over this range and has a minimum value of 0.25 mJ near the stoichiometric concentration. A sensitive flammable atmosphere obviously existed at the time of the incident since ignition occurred.
The above reasoning can be applied to electrostatic charge generated on the tanker driver. Here assumptions must be made since measurements were not undertaken on the driver involved in the incident or his clothing and footwear. The driver present at the tests had Texaco issued antistatic footwear and appropriate clothing and would not normally be considered a risk.
Conclusions Based on observations and measurements during this investigation and the analytical reasoning above, it is difficult to see how this ignition incident could be caused by an electrostatic discharge from the termination fitting. Other possible electrostatic causes which should be considered are electrostatic discharge from another source e.g. the driver or the aluminium cap. Can it be guaranteed that the driver was wearing issued antistatic footwear at the time of the incident? If the driver had been standing or kneeling on the manhole cover which is constructed from moulded plastic he may have become charged despite wearing antistatic footwear. Alternatively, the cap alone could become charged if it was placed on a charged cover and then transferred to the manhole in a gloved hand. Possible sources of ignition other than static electricity are impact sparks between the cap and the fitting and possibly thermite reaction. These sources could be investigated but at first glance seem unlikely. Although there is no technical evidence to support ignition by electrostatic spark in this instance it was recommended that the practice of grounding the transition fittings by means of a bonding wire (as undertaken in immediate response to this incident) be continued.
Graham Hearn is Director of Wolfson Electrostatics based at Southampton University. Wolfson Electrostatics specialises in the control of electrostatic hazards and can be contacted on tel: +44 2380 594995 fax: +44 2380 593015 or email: glh@soton.ac.uk