Recommended Practice for Lightning Protection of Aboveground Storage Tanks forFlammable or Combustible Liquids 1 Scope API RP 545, First Edition, Recommended Practice for Lightning Protection of Aboveground Storage Tanks forFlammable or Combustible Liquids, replaces the requirements of API 2003 regarding lightning protection forpreventing fires in storage tanks with flammable or combustible contents. This recommended practice (RP) providesguidance and information to assist owners/operators with lightning protection for tanks. This RP does not providecomplete protection for all possible lightning stroke occurrences. 1.1 Applicability This RP is applicable to tanks as described in API 650. 1.2 Application of Requirements to New and Existing Tanks The requirements of this RP shall apply to new or reconstructed tanks. The requirements may be applied to existingtanks at the discretion of the owner/operator. 2 Normative References The following referenced documents are indispensable for the application of this document. For dated references,only the edition cited applies. For undated references, the latest edition of the referenced document (including anyamendments) applies. API/EI Technical Report 545-A, Verification of lightning protection requirements for above ground hydrocarbonstorage tanks API Standard 650, Welded Tanks for Oil Storage API Standard 653, Tank Inspection, Repair, Alteration, and Reconstruction
API Standard 2003, Protection Against Ignitions Arising Out of Static, Lightning, and Stray Currents ASTM D3453 1, Standard Specification for Flexible Cellular Materials BS EN 14015 2, Specification for the design and manufacture of site built, vertical, cylindrical, flat-bottomed, aboveground, welded, steel tanks for the storage of liquids at ambient temperature and above EMMUA 159 3, Users' Guide to the Inspection, Maintenance and Repair of Aboveground Vertical Cylindrical SteelStorage Tanks NFPA 780 4, Standard for the Installation of Lightning Protection Systems SAE ARP 5412 5, Aircraft Lightning Environment and Related Test Waveforms 3 Terms and DefinitionsFor the purposes of this document, the following definitions apply. 3.1action integral The joule or ohmic heating energy dissipated per unit resistance at the lightning attachment point. The action integralis measured in A2s (amperes-squared seconds), which is the same as J 1 (joules per ohm) (Rakov and Uman,Lightning: Physics and Effects, p. 277). 3.2bonding An electrical connection between two electrically conductive objects that is intended to significantly reduce potentialdifferences. 3.3bypass conductor A conductive cable that provides a direct electrical connection between the tank shell and the tank floating roof. 3.4external floating roof tankEFRT
An aboveground tank with a floating roof, which has no fixed roof and has an open top. 3.5flash A complete discharge of the cell between the thundercloud and ground (as it applies to cloud-to-ground lightning)(Uman, The Lightning Discharge, p. 10). 3.6grounded (grounding) Connected (connecting) to ground or to a conductive body that extends the ground connection (NFPA 780). 3.7internal floating-roof tankIFRT An aboveground fixed roof tank with a floating roof inside the tank. 3.8release prevention barrier A release prevention barrier includes steel bottoms, synthetic materials, clay liners, and all other barriers orcombination of barriers placed in the bottom of or under an aboveground storage tank, which have the followingfunctions: a) preventing the escape of contaminated material, and b) containing or channeling released material for leak detection. 3.9shunt A short conductor that is electrically connected to the tank floating roof and contacts the tank shell. 3.10striking distance The distance over which the final breakdown of the initial lightning stroke occurs. 3.11stroke
One current component of a lightning flash. The number of strokes per flash is typically three to four, but may be aslow as one or as high as 30 (Uman, All About Lightning, p. 41). 4 Protection of Specific Types of Tanks 4.1 Fixed-roof Tanks (Metallic) and Tanks with Internal Floating RoofsFor fixed roof tanks (metallic cone or dome) and internal floating-roof tanks (IFRTs), there is a possibility of flammablevapors being present at atmospheric vents. If present, flammable vapors can be ignited by a lightning flash. Shunts or bypass conductors are not required for lightning protection. Bonding techniques to prevent static dischargebetween the floating roof and shell are addressed in API 650, Appendix H. Tanks handling low vapor pressures or in-service with properly maintained floating roofs with tight-fitting seals are notlikely to have flammable vapors at atmospheric vents unless it is being refilled from empty. In these cases, no furtherlightning protection is required (see Annex B). 4.2 External Floating Roof Tanks 4.2.1 Bonding Between Floating Roof and Shell 18.104.22.168 Shunts for Conduction 22.214.171.124.1 General Shunts are used for conduction of fast and intermediate duration components of lightning-stroke current. 126.96.36.199.2 Number and Placement The shunt to shell contact point shall be submerged at least 0.3 m (1 ft) below the surface of the liquid product. Theshunt shall have as short and direct a path as possible from the conductive floating roof to the tank shell. The shuntsshall be spaced at intervals no greater than 3 m (10 ft) around the perimeter of the
floating roof. When retrofittingexisting tanks with submerged shunts, the abovedeck shunts shall be removed. 188.8.131.52.3 Cross-sectional Area Minimum Width and Material The shunts shall consist of an austenitic stainless steel conductor of at least 20 mm2 (0.031 in.2) cross-sectional area,or of other material conductors of equivalent current-carrying capacity and corrosion resistance. The minimum widthof the shunt shall be 51 mm (2 in.). The shunts shall be of the minimum length necessary to permit the function of thefloating-roof seal assembly. The shunts shall be of the minimum length necessary to remain in contact with the shellduring the full horizontal and vertical design movement of the floating roof. 184.108.40.206.4 Durability The shunts and termination connections shall be of sufficient flexibility, crosssectional area, and corrosion resistanceto have a minimum service life of 30 years. 220.127.116.11 Bypass Conductors 18.104.22.168.1 General Bypass conductors are used for conduction of the intermediate and long duration component of lightning-strokecurrent. 22.214.171.124.2 Number, Length and Electrical Resistance The tank floating roof shall be bonded to the tank shell by direct electrical connection through an appropriate numberof bypass conductors. Each conductor, including connections, shall have a maximum end-to-end electrical resistanceof 0.03 . The bypass conductors shall be of the minimum length necessary to permit full movement of the floatingroof. Bypass conductors should be evenly spaced not more than every 30 m (100 ft) around the tank circumferencewith a minimum of two. 126.96.36.199.3 Durability
The bypass conductors and termination connections shall be positioned and of sufficient flexibility, cross-sectionalarea, and corrosion resistance to have a minimum service life of 30 years. 4.2.2 Parallel Conducting Paths (Seal Assembly from the Floating-roof Tank) Any non-fully submerged conductive seal assembly components including springs, scissor assemblies, sealmembranes, etc. shall be electrically insulated from the tank roof. The insulation level shall be rated 1 kV or greater. NOTE This allows any lightning discharge current from the floating roof to the tank shell to take the preferential path through theshunts and bypass conductors. 4.2.3 Insulation of Gauge or Guide Poles Any gauge or guide pole components or assemblies that penetrate the tank floating roof shall be electrically insulatedfrom the tank floating roof. The insulation level shall be rated 1 kV or greater. NOTE This allows any lightning discharge current from the floating roof to the tank shell to take the preferential path through theshunts and bypass conductors. 5 Metal Thickness Minimum metal thicknesses for tanks are provided in API 650 and API 653. Additional information is presented in EI/API 545-A, Verification of lightning protection requirements for above groundhydrocarbon storage tanks. 6 Inspection and Maintenance Requirements All bonding and grounding appurtenances shall be maintained and inspected in accordance with API 653. Belowdeckinspections shall coincide with API 653 outof-service inspections. General A.1 Phenomenon of Lightning and Secondary Effects on Tanks
A.1.1 Introduction This section summarizes the present knowledge on the lightning flash process and attachment mechanism. A.1.2 Lightning Principles Under fair weather conditions, there is normally a steady but weak vertical electric field at the earth s surface withvirtually no ground currents and a small distributed charge on the ground. Horizontal flat surfaces will have a veryuniform charge distribution, (i.e. a similar surface charge density everywhere). The highest surface charge occurs onthin pointed objects such as church spires, tops of aerials, tips of lightning rods, etc. Where the surface charge ishighest, the local electric field is the highest. Sharp-pointed, upward-facing items will tend to discharge a small currentinto the air, such as aerials and lightning air terminals. This will often be a silent, invisible discharge in the order of amicro amp. Electrical storms involve the relatively slow movement of heavily charged clouds. Charging mechanisms in the stormbuild up an electrostatic field over a large area across the base of the storm cloud. This field induces an oppositecharge on the surface of the earth beneath it. This induced ground charge flows along the surface of the earthbeneath the storm cloud at a relatively slow rate. The charging current flows are relatively small and cause nodamage. This charge differential is periodically neutralized almost instantaneously by a lightning stroke that collapsesthe field. At that time, a heavy ground current flows toward the lightning attachment point, equalizing local groundcharge distribution. The lightning process starts in the clouds, with a stepped leader descending to earth. The stepped leader oftenexhibits branching on its path to the ground as it attempts to find the best route to the ground. The path of the steppedleader is very irregular because of random variations in the local air conditions and other factors. When the steppedleader is within about 100 m (334 ft) or less from the tank (or ground), the electric field at ground level rises sharply,and the electric field on the highest items becomes great enough to launch an upward streamer towards the downcomingleader. In fact, two or more streamers may rise almost
simultaneously from ground objects (tanks, vents,trees, etc.) but only one usually is successful in making the connection to the downward leader (see Figure A.1). Thisis the usual mechanism by which a lightning stroke completes its path to the ground. A.1.3 Lightning Electrical Parameters A complete lightning discharge is called a flash. Each cloud-to-ground flash is composed of one or more lightningstrokes. Over 90 % of cloud-to-ground flashes are of negative polarity (Uman, The Lightning Discharge, p. 8). Atypical negative cloud-to-ground flash contains three to four strokes, but may have as many as 30 (Uman, All AboutLightning, p. 41). The currents in any one stroke can range from just a few kilo amperes (kA) to over 200 kA. The current in a typical negative cloud-to-ground stroke has several components, as listed below and as illustrated in A.1.4 Effects of Lightning A.1.4.1 General Tanks can be affected by both direct and in-direct lightning strokes. A.1.4.2 Effects of Direct Lightning Strokes It is standard nomenclature to name the point at which the lightning flash connects with the ground or structure as the attachment point. The attachment point for tanks will be at the highest vertical electric field regions which wouldinclude the tank rim, vents, hand rails, gauge poles, lights and other objects on the top of the shell or, for largediameter tanks, the fixed or floating roof itself. Lightning will not follow a single path down to ground. The stroke currentwill divide in proportion to the surge impedance of each available path. From the point of attachment, the current willflow as a sheet over all conducting paths. As the current spreads out over a large area, the surface charge isneutralized (see Figures A.3 a and A.3 b). Any discontinuities in the current paths may result in arcing across the gaps.
A.1.4.3 Effects of Indirect Lightning Strokes For a stroke adjacent to a tank, some current will flow over the outer skin of the shell across the fixed or floating roofand down to the ground on the other side of the shell (see Figure A.3 c). There would be much less energy in thedischarge currents moving across the tank as compared to a tank directly struck. As with direct strokes, anydiscontinuities in the current paths may result in arcing across the gaps. a) Top of Shell, b) Floating Roof, c) Ground Near a Floating-roof TankLightning flash togroundGround levelLightning flash toroofGround levelLightning flashCurrent flows downoutside of shell Ground level c) Current routes for flash to ground near a floating roof tank. The current spreads all around from thestrike attachment point, including to the tank, up and over the tank, and down the far side as shown by thetypical current flow lines and arrows. This current flow plan would only apply to the fast high current pulse. The continuing current would flow along the ground and the tank floor only. b) Current routes for flash to floating roof. Note that the fast high current pulse flows across the floatingroof in all directions to the rim seals and shunts, and then up and over the shell to ground. (Only if the roofis high is this a likely strike point.) a) Current routes for flash to top of shell. Note that the fast high current pulse flows down the inside of theshell and via the rim seals, and across the top of the floating roof. (Only two routes are shown; in practice,current flows all over the top of the roof and crosses the rim seal all around the perimeter of the roof.) 6 The examples in Annex A are merely examples for illustration purposes only. [Each company should develop its own approach.] They are not to be consideredexclusive or exhaustive in nature. API makes no warranties, express or implied for reliance on or any omissions from the information contained in this document.
Users of instructions should not rely exclusively on the information contained in this document. Sound business, scientific, engineering, and safety judgment shouldbe used in employing the information contained herein. 10 API RECOMMENDED PRACTICE 545 A.1.5 Sparking Sparking is the most likely cause of tank fires from lightning in external floating roof tanks (EFRTs) owing to the tendency for current from any stroke on the tank or closely nearby to drive the current across the floating roof, via the shunts or via any other metal making intentional or unintentional contact between the floating roof and the shell. Thermal and air-gap sparks should be considered as follows. a) A thermal spark is defined as a minute piece of incandescent material which has been ejected from somesparking site, usually a place where tens or hundreds of amps or more are passing through a very poor joint, suchas the contact point from a shunt on to the inner shell wall of an open floating roof tank, or from a poorly boltedflange joint, etc. The white-hot metal sparks falling from welding operation are examples of thermal sparks, whichare actually very small particles of metal, burning as they fly through the air. Usually, they are less effective asincendiary sources than air-gap sparks. b) An air-gap spark occurs in a location with a small gap between conducting items where the lightning creates avoltage large enough to cause electrical breakdown of the air or vapor/air mixture in the gap. c) Air gap sparks with energy above 0.2 mJ are sufficient to ignite product vapor/air mixtures if they are within theflammable mixture range. For mixtures that are not optimum, the energy requirement is higher, however, theenergy in lightning induced sparks is likely to be many orders of magnitude higher.
A poorly contacting rim-seal shunt on an EFRT is an example of where sparks might occur under lightning conditions. The small contact area and the presence of surface treatments or contaminants are conducive to sparking. If there isa nonconducting layer on the shell or the shunt, the spark will initially be an air-gap spark to break down the insulation,followed by current flow in a poorly contacting area resulting in thermal sparks. A.2 Grounding A.2.1 Flat-bottom tanks resting on the ground need not be grounded by the use of external grounding rods for thepurpose of lightning protection. Grounding for other purposes is not addressed by this document. A.2.2 The occurrences of incendiary sparks, rim-seal fires, etc., are not dependent on tank grounding resistance ortank dimensions. This is because the tank will inevitably have ground conductivity through its massive steel structurein contact with the ground. There will also be additional grounding through the many pipes and cables that connect toeach tank. A tank is considered adequately grounded if the tank bottom is resting on the ground or foundation. Thisapplies whether or not there is an elastomeric liner in or under the tank bottom. A.2.3 There is no known occurrence of lightning induced fires around the base of tanks, or underneath tanks, whichare the only places where inadequate grounding of the tank would result in sparking or other sources of ignition. Theinitial attachment process of a lightning stroke would be unaffected by tank ground resistance, since it will besufficiently low to permit the lightning leader attachment process to occur, even if the resistance to ground of the tankwere many tens of ohms. Lightning safety for tanks is not dependent on tank grounding. A.2.4 Tank grounding associated with the power frequency supply should conform to local electrical codes. A.3 Effect of Release Prevention Barrier Membrane on Grounding
Tank grounding is not an important contributing factor to the prevention of incendiary sparks or rim-seal fires, orexplosions in tanks. A release prevention barrier would be expected to have the effect of reducing the conductivitybetween the tank floor and the ground. However, the existence or absence of a membrane is not relevant to theprevention of lightning induced fires or explosions. 6 The exmples in Annex A are merely examples for illustration purposes only. [Each company should develop its own approach.] They are not to be consideredexclusive or exhaustive in nature. API makes no warranties, express or implied for reliance on or any omissions from the information contained in this document. Users of instructions should not rely exclusively on the information contained in this document. Sound business, scientific, engineering, and safety judgment shouldbe used in employing the information contained herein. Annex B (informative) Presence of Flammable VaporsIgnition cannot occur unless flammable vapors are present together with an oxygen concentration that places thosevapors within the flammable range. The Faraday Cage Effect serves to protect an internal floating roof from exposure. IFRTs have a good fire safety history, with four areas of fire vulnerability being: a) a landed floating roof (includes initial fill) exposes flammable vapors which are normally controlled by the floatingroof; b) improper operating practices such as an overfill situation (includes filling to the point where floating-roofs seals arenear the overflow vents) or storing product with a vapor pressure exceeding the limit and approachingatmospheric pressure (resulting in boiling) can expose flammable vapors normally contained by the seals;
c) mechanical failure (failing seals, lack of buoyancy, and other damage to the floating deck) can expose vaporsabove the deck; d) improper venting, which is venting other than as required by API 650, Appendix H. Annex C (informative) Different Seal Types Primary seals may be mechanical shoe seals, liquid-mounted rim-seals, or vapormounted rim-seals. Secondaryseals are vapor-mounted wipers mounted above any type of primary seal. 1) A mechanical shoe is a peripheral seal that utilizes a light-gauge metallic band as the sliding contact with theshell and a fabric seal to close the annular space between the metallic band and the rim of the floating roof deck. The band is typically formed as a series of sheets (shoes) that are overlapped or joined together to form a ringand held against the shell by a series of mechanical devices. Mini-shoe seals may be necessary for aluminumfloating roofs. 2) A liquid-mounted rim-seal is a resilient foam-filled or liquid-filled primary rimseal mounted in a position resultingin the bottom of the seal being normally in contact with the stored liquid surface. This seal may be a flexiblefoam (such as polyurethane foam in accordance with ASTM D3453) or liquid contained in a coated fabricenvelope. 3) A vapor-mounted rim-seal is a peripheral seal positioned such that it does not normally contact the surface ofthe stored liquid. Vapor-mounted peripheral seals may include, but are not limited to, resilient-filled seals (similarin design to liquidmounted rim-seals), and flexible-wiper seals. Flexible-wiper seal means a rim-seal utilizing ablade or tip of a flexible material (such as extruded rubber or synthetic rubber) with or without a reinforcing clothor mesh.
The sources of these definitions and descriptions are from API 650.
Few of us are so blasé or sophisticated as never to have felt our pulse quicken during a thunderstorm or to have never been entranced at sight of a lightning display. Primitive people ascribed many different explanations to the causes of lightning and thunder and at the dawn of the 21st Century for all of our education and sophistication the huge forces unleashed during a lightning strike are still capable of imbuing us with a deep sense of awe. The science of lightning protection is a fascinating field of study. Until recently however, it was a subject which was rarely treated in detail in engineering courses in our colleges and universities. The protection of buildings, and particularly electrical installations, against lightning induced damage is traditionally approached from two directions. Let s call them top down and bottom up. Following the top down approach on the one hand, a protection system of air terminations, down conductors and earth terminations may be considered usually based around the traditional Faraday Cage or Franklin Rod Constructions. Such a protection system aims to protect a building structurally and electrically by the installation of components which will provide a most attractive strike point, and path to earth, for any lightning strokes. The principal components of such a system are: - Air Terminals - Down Conductors - Earth Termination System It is important to remember, in this context, that lightning cannot be reliably predicted or prevented and traditional protection systems are designed to do neither.
The second - from the bottom - approach relates purely to electrical and electronic equipment ; consideration is given to the protection of alarm /instrument/ communications / electrical systems from the effects of lightning strikes and the associated phenomena. Protective measures for this second (or bottom up) approach, usually take the form of surge protection devices and a sensibly designed earthing system. In considering such protection systems, it is necessary to account for the possibility that damaging surges and voltage irregularities may be caused by remote lightning strikes and propagated over electrical lines. The provision of a lightning protection system for a complete building or installation is normally part of the design or construction responsibility of an electrical engineer as is the design of an earthing system. The implementation of the individual protection of components, devices or networks is the responsibility of the designer of the particular system in question, who may be a transmission, control or instrumentation engineer, an electrical engineer, a network systems administrator or may in fact be called by any title which is calculated to inspire respect and the payment of a high hourly rate. In this article, I propose to deal almost exclusively with the first approach as applied to industrial facilities. In Part I, I will attempt to give a brief overview of system design as usually practised in Ireland along with some comments. Then in Part II, three case studies will be presented which, hopefully, will serve to illuminate the points discussed, in the light of real life installations. The first of these has been included to illustrate that a protection system will not always be deemed necessary in this country depending on the type of facility being designed and the risk of lightning induced damage. The second case study will show the type of problems which are normally encountered during the design process and the third case study deals with an overseas project in which Australian standards were preferred over the British Standard which forms the basis of most lightning system design in this country.
The need for a Lightning Protection system in a new facility or building in a country like Ireland is debatable. The first step in designing a system should be a careful assessment of the risks of a lightning strike, the consequences of damage and the cost of protection. The relevant section in BS 6651 is an invaluable aid in this kind of assessment; however there will, of course, be factors individual to each project. Like any engineering tool, the assessment method in the BS is only that an engineering tool and not a substitute for thought or engineering judgement. Sometimes an engineer is asked to design a lightning protection system because it is an owner or an insurer requirement. While the demands of insurance companies will generally require absolute compliance, an Owner s brief may arise from a lack of understanding, on the part of the person preparing the design brief, of the nature of lightning protection and the benefits to be gained. A few points to bear in mind are:
<!--[if !supportLists]--><!--[endif]-->Locations in Ireland generally subject to only 2 4 lightning days per year. <!--[if !supportLists]--><!--[endif]-->Conventional lightning protection systems, do not keep lightning away, they merely provide a path to route it safely to earth some damage is to be expected even if such a system functions perfectly during a lightning strike. <!--[if !supportLists]--><!--[endif]-->Industrial facilities in Ireland are generally steel framed buildings with metal wall claddings. Except for high risk processes, for example, or for plants in which the initial (protection system) investment is negligible compared to the cost of damage repair and loss of production, or for areas in which an explosion hazard is to be expected, the benefits of installing a protective system in such a building are questionable.
Where a system is installed however, it is usually possible to reduce costs whilst maintaining functionality by integrating the lightning protection system with the building design - i.e. by utilising parts of the building structure as air terminations, down conductors or earth terminations. The design of a lightning protection system will normally be carried out in accordance with the recommendations of one or more of the internationally recognised design standards. Unlike the rather prosaic style which is the norm for engineering codes (no colourful cover shots or pictures of scantily clad models), lightning protection codes can be mysterious and wonderful documents awash with fascinating odds and ends of information (you can tell that I don t get out much). For example, British Standard (BS) 6651 contains a table showing the calculated risk of fatality associated with activities as diverse as smoking, mountaineering and driving a car. NFPA 78 from the US National Fire Protection Association has a section on personal lightning protection; including a list of good places to be during a thunderstorm (inside a large building) and bad places (inside a small tent). It also describes how you know that you are about to be struck by lightning (your hair will stand on end) and the recommended posture to be adopted - one which is suspiciously reminiscent of the old schoolboy joke about the posture to adopt during a nuclear attack - i.e. put your head between your knees and give the seat of your trousers a good-bye kiss (that is the posture recommended in the joke of course, not in the NFPA). The standard which is normally used in these islands is BS 6651. Many Irish designers are surprised to find that significant differences exist in the approach to lightning protection system design which is taken by standards and codes in other parts of the world (refer to Case Studies 2 and 3 in Part II of this article). If BS 6651 is adopted, then the design process will be broken down into the following steps: a) Risk assessment Is Protection required ?
This will take the considerations mentioned above into account and apply simple (and I do mean simple ) statistical techniques to determine whether the level of risk justifies the installation of a protective system. b) System layout etc. design and routing of termination networks, down conductors
The initial layout work for an industrial facility is normally quite simple once the major decisions such as the type of air termination network have been taken. The BS is well written and very easy to follow and it s recommendations can be interpreted to give guidance on all but very few of the situations which are encountered in the normal design process for industrial installations. The resulting layout drawings should show the roof termination system as well as the number and location of down-conductors and earth termination points. c) Specification and detailing Steps b and c are really one step (but I hope to get paid by the word). Along with the overall layout work it will be necessary to give detailed instructions as to how items such as earth termination points, connections, test links etc. are to be constructed. Detailed sketches will normally be prepared - either as part of an overall layout drawing or as an attachment to the specification documents.
A specification will be necessary to detail such items as the materials to be used, the final resistance to earth of the earth termination network and so on. This information must be presented in a form which treads the thin line between too much information (which merely repeats or conflicts with code recommendations) and too little information which makes the installation difficult to price and the design intent open to different interpretations thus making control of the installation quality more difficult.
Home Back to Top
Part II - Case Studies
In Part I of this article, I gave a brief overview of the design of lightning protection systems with particular reference to the design of industrial facilities here in Ireland. This month, I will present some case studies to try and illustrate some of the thrills and spills (precious few of either alas) which await the designer as he embarks on the design and construction process. 1) Case Studies Case Study 1: Project: Refurbishment of IDA advance factory to be refitted as a secondary pharmaceutical production facility (i.e. a finishing plant factory that makes finished drugs in the bottled, tableted, powdered or whatever form in which they will retail; as opposed to a primary or bulk plant which processes chemicals but does not carry out the production of the finished product ). Production was to be on a batch basis in a high grade clean room suite situated within the building. There was a large water treatment plant within the building while the boiler house, cooling tower and chiller were all external. Location: Building: Rural / industrial estate. Other Buildings well separated. Steel Framed, aluminium clad office / production / warehouse
building. 48m x 90 m x 7 m high. Discussion: On carrying out a preliminary assessment we found that the calculated risk factor was such that the BS recommended protection - but only just (i.e. the calculated risk was just on the border of the range for which protection was recommended). Our assessment of the cost of this system was of the order of IRP 20,000. My feeling, based on the estimated cost, the building structure and the nature of the process was that this would not be money well spent. Much of the potential cost was due to the fixing of the lightning protection tape to the roofing system which was proposed. This would be very time consuming. Furthermore, we were
informed that, for this type of roofing material, the fixing of the lightning protection components could only be carried out during a spell of fine weather. The above considerations were presented to our client and the requirement for a system was deleted from our scope. Case Study 2: Project: A greenfield development of a new production facility for the healthcare industry. Production was to be carried out in a sealed production line within a large low grade clean room. The process was to be continuous and if it were halted for any period of time, all of the product on the lines would have to be scrapped or reworked. The client regarded occasional stoppages as acceptable and on that basis the building power distribution was secondary selective (two transformers feeding a split switchboard with a normally open tie circuit breaker) with manual changeover no backup power system being installed for production or utility equipment. A substantial utilities suite consisting of boiler room, compressor room, chiller area and water treatment plant was located within the building while cooling towers and inert gas storage were external. Location: Rural/industrial estate. Other Buildings well separated. Building: Steel framed aluminium clad production/warehouse/office building with a large canteen/restaurant. The front face of the building was to be a glass curtain wall, side walls were to be of nonmetallic panelling for approximately 60% of the building length, while the rear of the building was to be clad almost to ground level with aluminium cladding. Discussion: A preliminary calculation indicated that protection was required.
For the reasons outlined in Case Study 1 above, our feeling was that a protection system was not necessary. We briefed our client on the factors involved and informed them that we felt a protection system was not required. In addition to the factors which I have mentioned already, the initial design brief for this facility was closely modelled on a similar production plant in the Southern United States which was subject to in excess of 90 lightning days per year obviously the new facility would be much less prone to lightning damage. However on consideration the client decided that a system was merited on the following grounds: The building would contain a number of through roof projections leading directly to the production clean room envelope. ANY loss of production would be expensive but a breach in the structure of the production envelope could lead to a prolonged shutdown which was to be avoided if at all possible. The total building investment was of the order of IRP 150 M including the production equipment. Serious lightning damage (for example due to a large fire) could lead to the loss of the complete investment, not to mention a shutdown which could halt production for over a year if the building were gutted. The existing plant in the US was protected by a system of vertical finials which were connected to a system of down-conductors. For the new system, we decided on a flat tape network in accordance with BS 6651. Even more so than with most electrical systems, the installation of lightning protection demands close coordination with the work carried out by non electrical trades - particularly if components of the building structure are connected to the system as outlined above. Any electrical system in construction work which requires a significant interface with other disciplines has the potential to cause major expense and lost time unless the coordination is properly managed. The work of the disciplines involved must be scheduled to allow the correct interfacing and then the installation contractors must be monitored so that they do not impede each others progress (note that when two groups of workers from different companies are working side by side and under pressure,
any breakages or damages will usually be the other group s fault; any lost time will usually be due to delays incurred by having to work around the other group and any responsibility for either of the above is virtually impossible to allocate. Moreover, claims of damage or impeded progress are very often based on fact. It is not unknown for installation workers to damage work already installed because it impedes the installation of their own equipment). On a well managed site, with professional installation contractors and good construction management, problems of this nature will rarely arise or - if they do, will be dealt with quickly and calmly. If the necessary coordination exercise is neglected then the whole issue can quickly deteriorate into a complete mess and will be a thorn in somebody s side regardless of how responsibility is allocated contractually. All that the designer can do is to make sure that firstly, the risk of such a situation arising is minimised and secondly, if it does arise, that specifications and scopes of work clearly define that the somebody in question (that is, the person in whose side the thorn finally sticks) is not himself (or herself or their client). In our system, we decided that at the sides and rear of the building we would use the structural columns as the down conductors and the reinforcing bars in the column feet as the earth electrodes. Using the reinforcing bars in this manner meant that a cable would have to be welded or bolted to the reinforcing bars, protected from future corrosion, protected from damage while the concrete was being poured and then protected from damage during the remainder of the construction process. Lastly if the cable was not installed or the connection was broken the responsible party would have to be made to pay for the substantial rework and expense which would be caused. To successfully achieve the above five in particular the last on a construction site is a feat that is difficult to fully appreciate unless one has actually been involved in the construction process. Likewise the installation of the roof network would require careful interfacing with the roofing contractor. Lastly the lightning protection system would have to be tested during construction.
The main design issues to be addressed were as follows:
i)Should we use a common termination network for the earthing and lightning protection systems? (i.e. connect them above ground and bring them to earth together as opposed to keeping the two networks segregated and bringing them to earth separately). ii)How would we design a system that could be tested periodically during the life of the plant (as is required by the relevant code and common engineering sense) iii)To what extent should metalwork on the building sides be bonded to the Lightning Protection System and how was this bonding to be achieved. i) The British Standard for Lightning Protection systems recommends that a common earth termination network be used for lightning protection and other earthing systems. The BS for earthing, on the other hand, only states that the two systems should be connected together at the main earthing terminal so as to prevent damage to the electrical systems in the event of a lightning strike. There was some concern that this could be taken to mean that separate termination networks were preferred. My preference is for a common network. This must be taken in the context of the position from which I view the design of electrical systems, i.e. as a designer of electrical power distribution and services systems rather than as a person who has to implement the design or the operation of sensitive communications or instrumentation systems. My reasoning may be summarised as follows: The lightning protection earth termination network is by it s nature extensive and easily accessible. The provision of a separate earth termination network is difficult to justify in terms of the time and materials involved.
The provision of so called clean earth networks (separate earthing systems for communications, control or instrumentation systems) is difficult to justify in many installations. It is often far more practical to maintain a system which seeks to obviate earth pollution through extensive bonding of different systems AND prevention of earth loops etc. than one in which coupling or connection between systems is to be avoided completely Moreover, if separate systems are provided and then taken to earth separately, dangerous voltages may appear between the different earth termination networks in the event of a lightning strike Naturally each installation must be considered as a separate case. Facilities which require the installation of equipment which is extremely sensitive to the effects of stray voltages and surges, or equipment which is intrinsically safe (a design philosophy for equipment which is to be installed in areas in which a gas or vapour potentially explosive atmosphere may exist), for example will demand special consideration. Our decision in this case was for a common termination network. ii) Another area in which slight differences in the approach taken by the two related standards occurs is with respect to testing of an earthing installation. BS 6651 (lightning protection systems) allows the use of the structural steelwork as part of the lightning protection system where practical and also permits the use of reinforcing bars in column bases as an earth electrode network. However it also stipulates that each earth electrode must be provided with a means of isolation for testing. These two concepts are clearly at odds
Any column in a system of structural columns, be it of steel or reinforced concrete, which is suitable for connection as part of an earth conductor network, will necessarily be in continuous electrical contact with the remainder of the building structure unless that structure is electrically non conductive. Such a conductor cannot be interrupted for testing in the sense that a single length of earth bar or tape leading to an earth pit can be interrupted.
The code does state that where structural columns are used as down conductors that sufficient points of test shall be provided to enable verification of the resistance to earth of the structure as a whole to be carried out, however concern was voiced that this could not be taken to mean any relaxation of the stipulation with regard to the testing of earth electrodes. BS 7430 (earthing) however lacks any inherent contradiction in this regard. It also admits the use of structural steelwork as part of the earthing system however it recognises the impracticability of attempting to interrupt or isolate part of the structural steel system for earthing and instead includes an alternative procedure for carrying out testing. Our solution to the need for testing of the system was to provide connection points on the building face to allow the connection of test instruments. We also provided for a test earth electrode in a remote part of the site (which was large enough that the test electrode could be sited 700 m away). In hindsight however, this last was not necessary as temporary electrodes can easily be used for this purpose. iii) A potentially serious issue was the bonding of extraneous metal on the building sides to the lightning protection system. The need for such bonding is easily demonstrated. When carrying the large currents with extremely rapid rise time which occur during a lightning strike, the elements of a lightning protection system can be raised to a high voltage level, despite being ultimately connected to earth. The most significant results of this are the danger of side flashing (arcing from the lightning protection system to adjacent, unconnected metalwork) and the danger that a person inadvertently connecting metalwork which is connected to the lightning protection system and adjacent, unconnected metalwork or earth, may be subjected to a dangerously high voltage so called touch voltages .
The accepted solutions to this problem are to rigorously bond all metalwork which may be a source of side flashing or dangerous potential differences, to the lightning protection system or to rigorously ensure, not only electrical isolation, but also segregation from the lightning protection system for all such metalwork. This sounds quite simple and often is. However, many installations include metallic items which fall into a grey area in which it is unclear if bonding is required. A commonly asked question, for example, is whether or not metallic window frames should be bonded to the lightning protection system (no - unless you have an EXTREMELY unusual window design) or water downpipes (sometimes). Some guidance to the determination of the need for bonding is given in the lightning code which includes a calculation method which acts as an aid to determining whether or not building metalwork should be bonded at each end, at one point, or not at all, to the protection system. Our grey area was occasioned by the discovery that the building design called for the use of metallic plaster stops around exterior wall panels. That is to say that the wall covering for much of the building exterior comprised a number of panels, each of which had a metal band around it s border. Our concern was to ensure that in the event of a lightning strike that these would not be the cause of side flashing off the lightning protection system with the attendant dangers of fire and electrocution and of course marking on the side of the building. Although this may seem remote, we could not say that the cross sectional area of the metallic strips was so small and the discontinuities between the plaster panels so large that such side flashing could not occur. We consulted the plaster stop supplier who said Well we never see this as a problem here in Germany.
(It should be noted that the Supplier s advice would be useful only up to a certain point. If they said:
Yes we always bond these items to the lightning protection system.
or better still
Yes we always bond these items to the lightning protection system and here s how.
Then we could investigate their solution and review it s applicability to our situation.
An answer to the effect that they did not consider this to be a problem told us nothing as we could not base decisions pertaining to lightning protection system bonding solely on the word of somebody who sells plaster panels for a living such people are not noted for being experts on electrical engineering matters.
In short, we were checking to see if they had encountered this problem before and could help us to find a solution and not to see if they could provide a solution for us).
After further study of the codes, we concluded that bonding was not required particularly in the light of the statement in Appendix 1 of BS 6651 to the effect that short isolated pieces of metal which are merely in fortuitous contact with the ground through the rain covered surface of the structure need not be bonded.
Case Back to Top
Extension to an existing oil refinery. The proposed extension was a large refinery development in it s own right and may be considered as a greenfield development.
Tropical (close to the equator).Coastal.
Building: A number of interconnected pipe racks, tank farms and process units / vessels, of metallic construction all located outdoors. Isolated control rooms and substation buildings were located around the site. The electrical system was secondary selective with automatic changeover. The highest voltage level on site was 132 kV. Sophisticated process control and shutdown systems were incorporated in the design giving a high degree of automation to the plant operations.
The location in question is subject to up to 140 lightning days per year (though to the residents of the surrounding area, it often seems that the true figure must be closer to 350 days). The existing site had suffered extensive damage to instrumentation, communications, fire alarm and electrical equipment due to lightning strikes, surge voltages and induced currents.
Lightning Protection was a very sensitive issue on this project.
The Project Specification (based on the Collection Volume method of Australian Standard NZS/AS 1768) was for a system of the enhancing type that would actively attract lightning strokes and safely conduct them to earth. This was to be achieved by the installation of air terminations which would actively respond to the presence of a lightning downleader in their vicinity by producing free electrons and thus attract downleaders from within a defined volume.
The specification for this system was a proprietary one - that is to say, it was a specification which could only be met by one company s equipment. Proprietary specifications are, naturally enough, the supplier s dream and the contractor s nightmare. An equipment supplier who has no competition in preparing a bid is under much less pressure to make it commercially attractive.
The project Turnkey (design, buy, build and commission) Engineering Contractor suggested an alternative system based on BS 6651 and JIS (Japanese Industrial Standard) A 4201 as modified by Oil Industry Standards and the Institute of Petroleum (IP) Code. It was felt that the Rolling Sphere method (as advocated by the BS) of determining the protected area around lightning protection system elements was more onerous than the method advocated by the Australian Standard. This would result in smaller protected areas being calculated and require more air terminations and down conductors, resulting in a higher degree of protection. It was also proposed to counter the high surge impedance of earth electrodes by embedding them in conductive concrete.
The Owner s view was that the BS was not really applicable for this sort of installation and was written more with buildings in mind. They felt also that the BS could be interpreted as to mean that very few downconductors would be required. Some of the information available on the existing installation indicated that the existing lightning protection system (which was to the same specification) was operating satisfactorily and that the equipment failures were primarily due to inadequate surge protection and poor earthingpractise (for example multiple earth cables, connected to separate termination networks, entering the same control rooms). Concern was voiced regarding the long term performance of conductive concrete.
It was decided to retain the original specification. The application of conductive concrete would be restricted to specific locations and deep driven electrodes would be used to lower the surge impedance.
It is impossible to deal properly with such a complex subject in a mere 4940 or so words. I hope though, that the case studies and discussions which I have presented will illustrate some of the design and construction issues which are faced in trying to wrap science about what is - in many ways - still the ART of Lightning Protection System design.