Early streamer Emission Lightning Protection · PDF fileEarly streamer Emission Lightning Protection Systems: An Overview Key Words: Air terminal, ... Corona discharge-A localized,

W F E A T U R E A R T I C L E

Early streamer Emission Lightning Protection Systems: An Overview

Key Words: Air terminal, early streamer emission, Franklin rod, lightning protection, lightning rod

arly streamer emission (ESF,) lightning protection sys- tenis are a relatively new approach to the perennial problem of lightning damage, and thcsc systems inay

hold promise for a more effective protection against lightning. However, tlie scientific and technical basis for this improved performance is far from certain and tlie efficacy of these technologies remains open to question. In this paper we examine the physical hasis for ESE devices and identify areas of controversy and gaps in our linowlcdge of lightning and lightning protection tliat need to be considcrcd in assessing ESE devices and in their future development.

The ESE devices considered here are liglitning attractors, and in that sense, their purpose is the same as that of conventional “lightning rods.” ESE devices, however, differ from conventional lightning rods i n that they are equipped in some fashion to increase thc efficiency of lightning attraction and thereby to cxtciid the effective range of protection over and above that of conventional lightning rods.

This article is hased on a compreliensivc hibliograpliy of ESE lightning protection systems tliat was preparccl at the re- quest of the National Fire Protection Research Fiiundatioii [I]; the reader is referred to that report for a critical analysis of the published literature. Most 11f tlie papers on ESE te rm- i i a l s have been published within tlic last 30 years, atid bibli- ographies covering the years hcfore 3 980 12, 31 reveal very little work directly concerned with ESE terminals.

Definitions As an aid ti1 tlie reader, ccimmonly used terms are defined

below. It should be cautioned, Ihowever, that tliere is little consensus 011 the meaning of these terms. For example, the terms “streamer” and “leader” are often iised iiiterchangc- ably. There are also cases where different words are used to denote the same phenomenon. For example, a “corona discharge” is sometimes called ii “point discharge” o r a “partial discharge.” The following definitions arc used in this report:

Lightning rod-A vertical conducting rod used to attract (or intercept) a lightning strike by prodncing a local ci i -

R. J. Van Brunt, T. 1. Nelson, and K. 1. Stric klett National Institute of Standards and Technology

ESE devices . . . differ from conventional lightning rods in that they are equipped in some fashion to increase the efficiency of lightning attraction and thereby extend the effective range of protection over and above that of conventional lightning rods.

lianccmeiit of the electric field strength in air. It is sometimes called a Franklin rod, conventional air terminal, or lightning conductor.

Early streamer emission air terminal-An air terminal equipped with a device that triggers the early initiation of the upward connecting streamer-leader discharge, when compared with a coiiveiitional air terminal under the same conditions.

Primary cloud-to-ground (CG) lightning stroke-The initial discharge between a thundercloud and ground that is generally associated with the propagation of a stepped Icadcr. Four types of primary strokes arc distinguished: I) downward propagating negative stroke from a cloud nega- tivcly charged relative to ground (often referred to as iiorinal lightning); 2) downward propagating positive strolc origi- nating from that part of a thuiidercloiid positively charged relative to ground; 3) upward propagating negative stroke; and 4) upward propagating positive stroke. The primary CG lightning stroke is also referred to as the initial stroke or as simply a lightning flash.

JanuaryiFebruary 2000 -Val, 16, No. 1 0683- 755410U1$1U.OUO2U0UiEEE 5

Stepped leader-lntciisc spark or plasma channel of fi- nite but variable length in air corresponding to the observed individual steps in a lightning stroke. This is considered to be a relatively higli-temperatore discharge stage heated by the passage of an electrical current pulse of high inagnitude [4-61.

Streamer-A narrow, lnighly directed, and self-propagating discharge i n air. A strcanicr dcvclops from an electron avalaticlie when the local space charge becomes of sufficient density to produce an electric-field strength coinparablc to or grcatcr than the external field. It is believed to propagate at a high velocity by the inechanisin ai photoionization in high-field regions produced ahead of the discharge. This is a relatively cold discharge phenomenon that can be the precursor to the forniation of a leader step [4-61.

Keturn stroke-This is a discharge that propagates upward from the ground (or lightning rod) in the cliannel formed by the primary downward stroke. An individual lightning event may cxhihit one or morc rcturn strokes [9, IO]. The return stroke should not he confused with the streamer-lcader initiated at a tcrtninal by tlie advancing primary lightning stroke.

striking distance-The distance covcred by the last leader step of a downward propagating primary lightning stroke i n making contact with a grouudcd ohjcct (lightning rod). This is sometimes called the final jump a i d it is expected to follow the path of an upward propagating streamer if such a streamer occnrs. 'This distance varies with type and intensity o f the lightning stroke.

Corona discharge-A localized, cold discharge in air that forins around ohjects such as sharp conducting points or wires that produce an eiilianceinent in electric-field strength sufficient to allow ionization growth. Corona is also helicvcd to form around Icadcr channels. Corona is an important

of space charge at grouud level and is sometimes called a point discharge, or partial discharge. Under soinc conditions it is the prccnrsor to streamer forination [I I].

Zone ofprotection-The volume surrounding or adjacent to a lightning protection system that is presumed to be sub- stantially iinnnune to direct lightningstrilies. In the case of air terminals, thiscould h e deiincdas thevolume in which an ac- ceptably high percentage o i lightning strokes will attach to the rod as opposed to other locations upon entering this vol- umc. A precise definition of the zone of protection and the methods to be used for its deterinitiation arc suhjccts of dc- hate [12-141

Cone of protection-A conic volume around a vertical lightning rod used to define a region of protcctioo. This is a cone whose height equals the height of the rod and whose base has a radius ccntcrctl at the rod and equal in length to the height (if the rod. The cone-of-protection concept is based on an electrostatic field [I5].

Rollingsphere method-A mcthud that enables identification of possible lightning attachment points by imaging a sphere of radius equal to the assumed striking distance that is rullcd over tlie exposed surface area surrounding a structure

(lightning rod) [14]. Points contactcd by the sphere on the surface identify possible lightning attachmetit locations. It should be noted that, although the method identifies possi- h l e attachnient points, it does not give iniormation ahout the probability of attachment to these points [I 61.

1,ightningattmction efficiency-A measure of the proh- ability that a lightning stroke will attach to an air terminal i f it enters its zone of protection, e.g., a 90% cfficicncy iin- plics that YO"h of all strokes that elites the zone will attach, independent of parameters such as current and angle of approach. There seems to he no clcar consensus on this deiinition; howcvcr, the above definition will be used in this report.

Onset field-The electric-field strength above which ioii- ization growth (clcctron-avalanclie formation) is possible in air. This is approximately 2.6 MVhn in dry air at standard temperature and pressure [17].

Time lag-The time hctwccn when the field strength first exceeds the onset iield and when a discharge is initiated. A distinction is usually made between the statistical time lag (t,) corresponding to time of electsoti-avalanche initiation and the formative time lag (T~) corrrsponding tn the time for complete discharge (streamer or leader) formation. Usually (zf) is much less than (zs) [ 171.

Space charge-Density of charged particles (ions) in air tliat modifies the local electric field. The cxistcncc of space charge can have a significant influcncc 011 liglitning propagation.

Thundercloud-A cluud cuntaining a charge density sufficiently high to allow iormation o i a lightning strokc.

Cloud-to-cloud (CC) lightning stroke-A lightning stroke between thunderclouds that may or may not be related to a cloud-to-ground stroke.

Iightning dissipater array-A system that supposedly rc- pels or diverts lightning by fortnation of space charge such as from corona discharge generated by an array of sharp metal rods [20, 211.

Electron avalanche-An electron multiplication process due to electron-impact ionization of gas molecules. This is the initial stage i n the development of an electrical discharge in air, e.g., a corona or streamer.

Discussion Characteristics of Lightning

In the discussion that follows, we highlight some of the properties o i lightning to which particnlar attention should be paid in the evaluation of liglitiiitig-attractor-type protection systems. A cotnpletc trcatmcnt of the state-of-knowledge of lightning can he found in other publishcd reviews 13, 22-26].

Perhaps the most significant property of lightning that should be kept i n mind is its complex, stochastic, and fractal character. These properties make it impossible to predict precisely how a lightning stroke will develop. Indeed, the behavior of lightning phenomena is best descrihrd in statistical terms, e.g., by giving probability distributions for stepped

6 IEEE Electrical Insulation Magazine

leader lengths, striking distances, currents, number of return strokes, angle of approach, etc. as are often found in the lit- crature [27-361. It may be noted that evcii low-level corona discharge phenomena can exhibit complex stochastic and niulti-modal behavior that is not easily predicted or undcr- stood 111, 37, 3x1.

Two types of cloud-to-ground lightning discharges arc identified, namely, positive and negative, where tlie sign re- fers to tlie charge at the base of the cloud relative to thc charge at ground level. Upward propagating dischargcu, which apparently originate at groutid level, of both polarities are also observed. In most locations, ticgative downward strokes are by far the prcdoininant form of lightning. How- ever, recent surveys show that the proportion of positive strokes can be greater than 25% in some locations, especially in mountainous areas or high latitudes in the northern hemi- sphere where thunderstorms associated with frontal zoiics predominate (28, 39, 401. The occurrence of positive lightning also varies significautly with season, generally being most common in winter 132, 39, 411.

In general, the statistical distributions of lightning characteristics vary with terrain, altitude, latitude, a d time of year [42]. These variations need to be considered in evaluating or designing a lightning protection system for a particular location. Evidence for significant variability in the behavior of lightning is fouud not only from geographical survcys [40, 431, but also from nmnerous observations and recordings at sites frcquetitly hit by lightning, such as the CN tower in To- ronto, Canada [44], the Empire State Building in NcwYork [28,3 1,34,4.5], Mount San Salvatore near Lugano, Switzer- land [27,46], and otlicr locations [30,47,48]. Observations at such locations reveal imiisnal and uiiexpectcd behavior of lightning propagation [49]. Although it is known, for example, that lightning often propagatcs horizontally with respect to ground level within a cloud [SO], there are also unexpected recordings of nearly horizontal propagation over considerable distances at sub-cloud levels [46, 51-54]. The phenomenon of “hall lightning” [SS-SS1 inay corre- spond, in some instances, to a type of horizontally propagating lightning near ground level.

Lightning is gcnerally helievcd to propagate in the atino- sphcrc hy a streamer-leader mcchauism [4, 5, 35, 59-66]. Leaders arc highly conductive, luminous plasma chamiels that typically vary in length hetween 3 in and 200 in. The initial lightning stroke is composed of conuccted leaders or leader steps. Tlie avcrage velocity of propagation of a stepped leader gcnerally fits a lug-normal distribution and typically lies within the range of 1 .0 x 10’ i d s to 2.7 x 3 O6 m/s [36, 671, and the electric charge deposited in the leadcr channel is within tlie range of 3 “C tu 20 “C.

Because of the high concentration of charge in the channel, tlie electric-field strength produced by the leader can he sufficient to allow elcctrical breakdown (ionization) of the air at distances greater than 10 in from the tip of the leader. Corona-type discharges develop i n tlie intensc field region at the end of tlie leader from which fast streamers cniaiiate aod

prepare the path for the next leader step [4,61,64]. When a leader approaches ground level, it caii, given a sufficiently high charge concentration in tlie channel, cause the field strength at groundcd conductors such as a lightning rod to hccoiiie great enough for tlie initiation and developnicnt of a discharge at the grounded site. If a discharge is initiated at ground level, it is likely to propagate upward in the high-ficld region produced by the leader channel [68]. This is usually the proccss hy which the initial lightning stroke completes its path to ground. Tlie existence of upward propagating discharges that connect with a downward propagating leader has bccn clearly established from observations in tlie laboratory of discharges generated under impulse voltage conditions in long air gaps [@, 701 as well as from obscr- vations of natural lightning 146, 68,711. It should be noted, howcver, that upward coniiecting streamers are iiot always observed during a liglmiing strike to ground [25, 711. Con- sidering tlie probabilistic nature of the streamer initiation process, it is not surprising that thc upward streamer inay not always occur, or may be too weak to observe. Moreover, thc advancing leader caii initiate more than one upward streatncr at different ground-level locations [25, 721, hut only uue is likely to connect.

The probability that an upward streamer will be initiated at agroundsite increases with thelocalelectric-fieldstrength produced by tlie leader. The instantaneous field produced by the lcader in turn increases with the instantaneous charge and therefore tlic current in the leader channel at any given time. It can thus be expected, as observed, that for a fixed striking distance from a groundcd object, tlie probability that ai1 upward strcamer will be launched should increase with increasing current in the final leader step of the lightning stroke 1681. Also, the mean striking distance to a grounded object should increase with increasing stroke current [22, 731.

A typical lightning stroke exhibits a complex branching structure [3,45,48]. This branching comes about in part hecause of the possihility that, at the end of a stepped leader channel, more than otic path inay be prepared by different streamers [63]. If this happens, tlie subsequent leader step caii divide and propagate simultaneously i n two or more different directions. Only one of tlie leader channel branches is likely to connect to ground. Cases have been recorded, however, where two or more stepped leader branches simultaneously propagate to ground. In such cases, questions arise about the inutual interaction between branches and the niin- imum likely branch separation. The theory of lightning is nut sufficiently advanced at the present time to account for branching behavior and the general fractal dimensions of lightning. The space-charge distribution in the atmosphere below the cloud is also likely to influence the branching characteristic [131. It is only relatively recently that fractal models of electrical-discharge phenomena have been introduced that deal with the branching characteristics [30, 741.

There is still much that we do not understand about the hehavior of lightning, and the physics of lightning remains a

JanuaryIFebruary 2000 -Val, 16, No. 1 7

topic of intensive scientific investigation at research laboratories arouud the world [75-781. Little is known, for example, ahout the mccliaiiisms hy which a liglitning discharge is initiated within a thundercloud [79, 801. The average electric-field strength in a cloud generally lies far below the breakdown strength of air. The initiation of lightning is possibly associated with local cnlianccnicnts in the charge density within tlie cloud. In any case, there is reason to helieve from observations of liglitning propagation within clouds 154, 8 1 J that a cloud from which liglitniiig originates cannot be treated as a uniformly charged equipotential region. Moreover, it has been argued that objects at ground lcvcl cannot influcncc the triggering of lightning in a cloud [82, 831. l‘his is a reasonable expectation because tlie region of a cloud from which liglituiiig originates is likely to he at an altitude of 3 kin to 4 kin above ground level, which is one to two orders of magnitude higher than most grounded objects. Thus, any claims made about the ability of a gromid-based lightning protection system to trigger or attract lightnitig strikes from a cloud should he viewed with skepticism. An exception, of course, is the use of grouud-lauiiclied ohjects such as aircraft or rockets that are known to trigger lightning [84, 851.

Aiiothcr area of active lightning research that deserves special attention is that concerned with tlie influence of local space charge on the path of a lightning stroke. It has already been shown from laboratory simulations [54, 86, 871 that discharge propagation can he affected significantly by the presence of spacc charge associatcd with iuns or charged particles. Space-charge effects are expected to he significant within a tliimdercloud, hut may also he important at ground level where ions can he produced by corona or point discharges that occur i n advance of a liglitniiig strike 123, 82, 88-91]. The presence of space charge could influence the performance of a lightning protection system and may ex- plain tlie occurrence of horizontal or other unusual forms of lightning 11.31. Kcscarch is nccdcd on factors that influence the electric-field distribution at ground level under a thundercloud 1x1, 92-95],

ESE Air Terminals An early strcaincr cmission air teriiiiiial differs from a

convcntional air teriiiiiial or liglituiiig rod in that it is equipped with a device that supposedly eiihanccs the probability of initiating an upward propagating discharge (streamer) to connect with tlie downward propagating leader of a lightning stroke [96J. I’rcsumably, this enhancement applies for both polarities of iiatural lightning, although discussion of the polarity dependence of ESE systems is noticcahly ahscnt from the literature. An ESE termiiial can often h e distinguished from an ordinary lightning rod by the presence of a small object iiear the top of the rod that serves as a discharge trigger [Y 71. Tlie geometrical configuration of the rod tip for ESE rods can also be more complex than that of a conventional rod [97].

There arc diffcrcnt types or designs for ESE termiiials. A coininon feature in their operation is tlic use of a discharge triggering device tu incrcasc the probability for initiating a streamer discharge at or near the rod tip upon the approach of a descending leader. Contrary to some misconceptions, ESE terminals do iiot significantly incrcasc the conductivity of air at a distance greater than 10 cm beyond the tip of the rod [98, 991. Possible exceptions are systems that utilize in- tense laser beams to “guide” a leader discharge. However, such clcviccs arc considered to be experimental and are not presently used for practical lightning protection. Tlie main attracting effect of an ESE terminal is undoubtedly due to the metal coilductor itself that introduces a significant enhancement vf tlie electric-field streiigth which, in turn, increases tlie rate of ionization in the air around the rod above that at other nearby locations. In s o doing, it in ity of discharge initiation and perhaps also tlie speed with wliicli the discharge can propagate compared to surrounding areas that arc at a lower field. A tall lightning conductor can therefore compete more favorably in attracting a lightning discharge than conductors in the same vicinity of smaller height [IOO, 1011. Like conventional tcrniinals, tlie attraction efficiency of an ESE terminal should increase with its height abvvc the ground up to a limit determined by the inaxitnuin striking distance.

Three general types of ESE devices have been identified, namely: 1) rods to which a radioactive source is attached, also referred to as ionizing o r radioactive air terminals 1102- 104); 2) rods equipped with an electrical triggering device [97, 10.51; atid .?I) systems that use laser beams [I 06-1 091. Of these otily the first twv types are currently in use. The third type is still under development and its effectiveness, although promising, has iiot yet heen demonstrated outside of the laboratory.

Tlic most widely used, and perhaps most controversial ESE device is that equipped with a radioactive sviirce positioned near tlie top v i tlie terminal 1102, 110, 1 I I]. The radioactive materials employed are weak alpha particle emitters with relatively long lifetimes such as L4‘Am (half-life of 433 years), which is also used in some smoke detectors. Other types of radioactive materials have also hcc i i used, including 210~’o, 226Ra, ”I<r, and 6oCo 11121. The products from radioactive decay of these materials ionize the air in the immediate vicinity of tlie terminal, typically within a radius of 1 ci i i to 3 cni. The ion-pair production rate can be as high as 1O1’/s. It has been argued that, outside of a small region near the terminal, the impai r forniation rate in the atmosphere from the radioactive sourcc will fall sigiiificautly below the rate from natural hackgroutid radiation [99]. Evidence that radioactive air terminals are superior to conventional Franliliii rods has heen reported based on interpretation of results from outdoor tests usingirnpulse hreakdowii of rod-planc gaps L113-117]. How- ever, many questions have been raised in the literature both about the effectiveness of radioactive air termiiials 1.12, 16, 90, 98, 99, 118-1291 a d about the potential hazards they pose due to pussihle human exposure to harinful radiation


11 12, 129-1351, Indccd their use is proliihited in some COLIII-

tries [lOS]. In recent years, ESE teriiiiiials that use electrical trigger-

ing devices have been introduced. In principle, their purpose is tlie same as a radioactive source, aiid there is uo reason to believe that electrically triggered systems could not be at least as effective as a radioactive air tcrminal. They offer tlie potential advantage of more control over ion production at the tip of the terminal and thus would appear to be more effective i n avoiding unwanted colitinnous ciirona discliargc formation in advance of a lightning strike. They also avoid tlie healtli and cnvironmcntal issues that arc associated with radioactive devices. Nevertheless, very little information could he fonnd about electrically triggered ESE systems in tlie archival literature. Most (if the discussiiin ahout these dcvices is confined to patent documents ['I 051 aiid to relatively recent conference papers [97,136-1401. The availablc information is usually both of a preliminary nature and lacking in technical details. Consequently it is iinpossiblc at this time to malie a complete, independent assessnient of tlie performance of ESE terminals that employ electrical triggering.

From tlie limited infiirtnation published about such devices, it appears that they empliiy a detector that s c u m the approach of a downward propagating leader by producing an electrical signal proportional either to the electric field or the rate-of-change of tlie electric field produced by the ap- proacliiiig leader [Y7, 10.51. Wlicn the output signal of the detector reaches a certain level, it triggers n circuit that then applies a fast, high-voltage pulse or pulses either directly to the rod o r to a spark gap electrode arrangement positioned at the top of the rtid. The application of the electrical pulses enlintices tlie field enough to create a local discharge or ionization at the most opportune time to initiate :in upward propagating streamer. Thus, unlilie a rod equipped with a radioactive sonrce that causes continuous ionization in the snr- rounding air, the electrically triggered ESE device produccs ionization only during a brief pcriod prior to the lightning strike. Information about the duration and extent of this ioii- izatioii conld not be found.

There arc other types of recently devckiped ESE terminals that utilize tlie piezoelectric effect for electrical triggering. Although this type of terminal is now available commercially, only scant incntion of it is made in tlie archival literature [105, 1411. From an examination of the literature, we decided that tlicrc is insufficient published information about the operating principles and performancc of the piezoelectric devices to enahlc an independent ap- praisal of their performancc. It is presumed that tlie function of a picziielectric trigger is the same as other ESE dcvices, namely, to enhance ionization and thereby increase tlie probability (reduce the time lag) for initiation of an upward propagating streamer.

As mentioned above, lightning protcction systems that utilize laser beams to trigger and/or guide a lightning discharge, o r to assist in launching an upward streamer arc coli- sidcrcd to he experimental. At tlie present time, there are no

known commercial systems tliat employ lasers. Laboratory stndies have shown, however, that laser heains can he effective in initiating and controlling tlic path of an electrical discharge in long air gaps 1108, 142-158].

The comparative advantages and disadvantages of tlie different types of ESE devices i n t e r n i s of their cost, size, reliability, etc. have not been examined in this investigation. The fact that therc arc competitive ESE devices available with differing configurations aiid operating principles should be of concern in proposing uew standards for liglitning protcctioii systems. It is douhtfol that tlie efficiency aiid reliability of performaiicc is tlie same for all types of ESE terminals.

In the following three sectious tlic physical bases for liSE devices arc examined and issues rclatcd to tlie validation or verification of ESE device pcrforoiaiice are discussed.

Physical Eases for ESE Air Terminals A complete assessnieiit of ESE air-terminal pcrformancc

rcqnires an understanding of the basic physical mcclianisuiis responsible fo i - their operation. It is not clear from an examination of the relevant literature that tlici-e is a complete aiid universally agreed upon understanding of liow or why tliese devices work. In the case of ESE terminals that use radioac- tivc S O U I - ~ C S , argunients arc made in tlie literature that they simply do not work, or arc at best relatively ineffcctive [16,

I 18, 120, 222, 124-'129]. These negative opioious arc USII-

ally based on a lack of cvidcncc that could be found in tlic literature about tlie performance of radioactive terminals, and arc seldom hased on results of independent tests. Reliable qiiantitative data ahout the rclativc perfoi-rnancc of ESE vel--

sus coiiveiitioiial dcviccs under relevant conditions arc defi- nitely lacking. However, die lackof data does not necessarily prove that ESE devices do not work. 'The questions that should be asked are, how do they work and how much hetter do they work, or could they work than couventional ternii- tials, i.e., what is the gain, if anything, in lightning attraction efficiency? Indisputablc results were not found to auswci- these questions.

Taking into consideration what is presently known from lahoratory studies of relevant electrical-discharge phcnoiii- cna, reasonable speculation is possible about the iinpiirtant processes that can account fur ESE device operation. Uiifor- tunatcly one must resort tii speculation or extrapolations from lahoratory scalc cxpcriments because tliere is a dearth of detailed informatior from iibservations made during lightning strikes in tlic natural environmcnt.

The one characteristic that appears conmion to all types of ESE devices is that they enhance ionization of the air i n tlie immediate vicinity of tlic terminal tip prior to an approacli- ing lightning stroke. This additional ionizatiou presumably enhances the probability that a n upward propagating streamer will he lauuched froiii tlie teriniiial tip. Alcgitimatc qnestion that could then be asked is, liow does this additional ionization act to enhance streamer formation?

JanuaryiFebruary 2000 -Vol. 16, No 1 9

The aiiswer to this question is not obvious but would appear to he found from a consideration of the time lug to clcc- trical breakdown. This topic has been the subject of numerous laboratory investigations [19]. In order for an electrical discharge (streamer) to he initiated after a rapidly rising voltage has been applied-for example, to a point-sphere or sphere-sphere electrode gap in air-so that the electric-field strength exceeds the breakdown streugtli of air, there must be at least one fsee electrou available to initiate the electron avalanche process, which is tlie precursor to streamer formation [4,7, 19,159,1601. In tlie case ofaposi- tivc point electrode, which approximates tlie conditions of a normal negative liglitning stroke, the initial electron rclcasc mechanism is thought primarily to be collisional electron detachment from negative ions 119, 1611. The rate of collisional detachment depeiitls hotli on the anion species and on the strength of the electric field in which it moves.

In the case of air, the types of negative ions that can be formed depend significantly on water-vapor content (liumidity), which appears to account for tlie observed large dif- ference in laboratory measured time lags for dry arid humid air under positive impulse conditions 1162-1661. From laboratory experiments on positive impulse breakdown in air, it is usually found that measured time lags exhibit a pro- nounced decrease with increasing humidity, i.e., the discharge initiation probability is enhanced hy the presence of water vapor [19, 1671. Negative ions in the atmosphere are formed by attaclunent of low-energy electrons to elec- tro-negative gas molecules such as 0, and H,O. Initially formed negative ions such as O ~ , O ~ , and OH^, can undergo transformations into other types of negative ions such as 0, and OH~.H,O through a complex sequence of ion-molecule reactions 1111.

The presence of negative ions in an electrode gap prior to the application of an impulse voltage (simulating an approaching leader) does not guarantee that a discharge will be initiated. For example, if the rate of voltage rise is too slow, the ions inay simply be swept out of the gap before undergo- ing detachment. Moreover, the field strength at which a negative ion can detach an electron may, depending 011 tlie type of ion, lie above or below the breakdown field stscngth of air and this will determine its effectiveness in iuitiating a discharge. If tlie ESE device helps ensure the presence of negative ions uear tlie terminal during tlie approach of a lightning stroke, then it could he effective in reducing the time lag for streamer initiation. However, its effectiveiiess must he inea- sured against naturally occurring time lags and might depend significantly on such conditions as relative humidity and total charge (strength) of the oncoming Icadcr in the lightning stroke.

In the case of a negative poiut electrode, as occurs for positive lightning, the mechanism for discharge initiation can bc quite different. Near a negative point, detaclimcnt of negative ions may still play a role; however, these ions will be forced to move into tlie lowes field region away from the electrode tip where electron detachment by collision with air

nioleculcs becomes less probable. For negative points, electrons can also be released by collision of positive ions with the electrode surface. Tlie effectiveness of preleader ionization in enhancing streamer initiation at the tip of an air ter- iiiiual can lie expected, therefore, to depend on polarity. The extent to which the probability of streamer initiation depends un polarity for a given ESE or conventional terminal is generally not known, or at least there is no evidence in the archival literature that the effect of polarity has been investi- gated thoroughly.

The presence of ionization at tlie terminal tip in advance of a lightning stroke can also act to undermine tlie effectiveness of an ESE device if this ionization can occur under high enough field strength to allow formation of a corona discharge 142, 'I 13,160,168,1691. Oiiceacoronaforiiisitpro- duces orders of magnitude more ions than can be generated, for example. from alpha particle emission from a radioactive source. Tlie presence of ion space charge can significantly rc- duce the electric-field strength near the top of ai1 air tcriniual and inay thereby act to inhibit streamer initiation 1.37, 1701. The possibility of corona formation depends on factors such as the geometry of the tcrminal[171]. It lias also been shown conclusively by experiments performed in the atmosphere that the intensity of a corom discharge aiid tlie density of space charge associated with it depends significantly on local wind velocity 194, 1721.

Laboratory experiments performed to determine tlie influence of radiation 011 the initiation of air discharges in large sphere-plane gaps have shown that the presence of radiation

ascs the likelihood of discharge formation for impulse voltages with steep wavefroiits, but decreases the likelihood for breakdown at longer wavefronts (1 ps compared to 180 ps) [173]. For steep wavefronts the breakdown voltage is therefore effectively lower than it is for longer wavefroiits. This experiment seems to show that the presence of radiation enhances discharge initiation provided there is insuffi- cieut time for corona space-charge formation. However, the effect of the radiation was, in either case, relatively small and

and role of corona genesated space charge were not quantified or even clearly identified. One of the conclusions given in this work is that the major effect of tlie radiation and corresponding ionization of tlie air is to eliminate very long time lags to spark-over. It should be kept in mind that time lag is a statistical variable, aiid for a given wcll-dc- fined set of discharge gap conditions there will exist a distrihution in time lags that can be dctcrmined experimentally [19, 1611.

It should also he noted that corona discharge formation is involied to account for tlie effectiveness of supposed lightning dissipaters [20, 12, 121, 1631. In this case, coroua discharges presumably form at a multitude of sharp conductors positioned around the area to be protected and thereby produce enough space charge to reduce the electric field and deflect tlie path of oncoming liglitiiing. The forination of corona can depend siguificantly on tlie water vapor content of tlie air arouud the terminal [164, 1741. Tlie

IEEE Electrical Insulation Magazine

issue of corona formation is also central to the debate about the most desirable shape for the end of a conventiunal lightning rod [21,83, 3 75-1 781. Blunt rods are reported to pcr- form hctter than sharp rods i n attracting lightning, supposedly hecause corona forms less readily around hlunt rods than around sharp rods 11781. The issue of corona for- niatioii is also relevant to the operation of ESE devices and is presumably a factor that is considered in the design of ESE terminals. Details of liow ESE devices are designed to avoid corona formatior prior to a lightning strike are not discussed iu the archival literature.

111 additiou to iou formation that occurs during discharge activity near the tip of a lightuing rod, clcctroti and inn collisions with atmospheric molecules can form relatively long-lived metastable excited neutral species such as the a‘Ag electronic state of tlic oxygen rnolccule or the A’Z;, electronic state of tlie nitrogen molecule. Additionally, vibrationally excited metastable molecular species are produced in a gas discharge. The presence of these nietastablc species can have a significant influence oii strcamcl--discharge propagation because they are more readily ionized thau air niolecules ill the ground state and hecause they can supply energy to electrons hy super-elastic collisions, the overall effect of which is to providc a path of lowcr resis- tance to an oncoming discharge [17Y-1 SI]. The qucnching of metastable species through collisions with other molecules or with surfaces can also he a source ci f discharge initiating electrons, i.e., their presence under some conditions might he effective in enhancing the probability of discharge inception 1381.

Uiililie ions, ncutral metastable spccies do not contribute to modification of the local electric-field strength and their motion is also not sigiiificaiitly influenced hy tlie prcseuce of a field. l‘hey tend to diffuse away f r i m thcir point of origin, and their effectiveness in modifying a discharge path and their range of influence dcpcnds ou their excitation energy and density distribution at any given time. The density of metastable species depends, in turn, on the relative rates of formation, quenching, and diffusion of these species. Al- though the influence of metastable species on discharge de- velopinent has bccn estahlislied from laboratory iuvesti- gations [180], considerahly less is liiiowii ahinit the dynamics and interactions of these species in a discharge compared to what is linown about ious. In particular, very little is ktmwn about how they contribute to lightning discharge initiation or propagation under relevant atmospheric conditions. As with negative ions, tlie metastable content of the air around a lightning teriniiial will be affected hy relative humidity and general air contamination. ‘I‘lie influence of metastable species should not extend significantly heyond the elid of a lightning rod. Their role, if anything, will be to enhance initial development of a streamer at the rod tip.

In summary, it would appear that enhancement c i f upward streamer initiation from an F,SE terminal (compared to a conventional tcrminal) has a plausible physical basis. How- ever, it would also appear that a conipletc and universally ac-

cepted undcrstaudiiig of how a11 FSF, dcviccs work has iiut yet heeii achieved, and it can he argued that a better uuder- standing is nccdcd to make ineaningful quantitative conipar- i s o t i s hctwccn tlie performances o f ESE and conventional devices. To reach such au understanding it will undouhtedly be necessary to address ntiineroti~ basic questions such a s :

1 , What are the prcdominant streamer initiation mecha- nisms under different conditions of polarity, atmospheric liumidity, air contamination, atid terminal geometry?

2. What are the relative roles c i f ious, electrons, and inetastahle species or tlic development and propagation of a streamer discharge from a terminal for different conditions?

3 . What is the likelihood of corona formation around a terminal aiid how will thc presence of corona affect the ability of the terminal to launch a streamer upon approach of a lightning stroke?

4. In the case of radioactive terminals, what is tlie dcpend- eticc of the streamer initiation probability on the intensity and type of radiation source!

5 . In the case of electrically triggered devices, how docs the streamer initiation probability depend on the timing and inagnitude of tlie electrically triggered spark?

6. Also for electrically triggered devices, liow reliable is the field sensor that controls die triggering, and ani its performance bc affected by local spacc charge? Attempts to find answers to questions like these arc the focus of much ongoing cxperiinental and theoretical research, not only on lightning, hut also on electrical discharge plicnom- ciia in general.

Validation of ESE System Performance Three general methods have hecn used to evaluate and

test the performance of lightning protection systems, namely: 1) small-scale laboratory or outdoor tests in which liglitning, or the effects of lightning, are siiiinlated by apply- ing high-voltage impulses to widely separated electrodes; 2) theoretical simulations of lightning strokes that predict propagatiou behavior and striking distance; aiid 3 ) outdoor tests involving observations of artificially triggered or uatil- rally occurring cloud-to-ground lightning strikcs. 111 this scc- tion we briefly examine tlic advantages, disadvantages and issues that havc heen raised coiiccrning the use and validity of these mcthods.

LABORATORY AND SMALL-SCALE TESTS Considerable insight has bccn gained ahout the physical

nature of lightning from laboratory-scale studies of electrical breakdown and spark forillation in “long” air gaps, with typical gap spaciiig of 2 111 to 1.5 in [69, .136, 166, 173, 182-188]. Iring air gaps have also bccn used to test the performance of lightning rods, including ESE devices, both iu enclosed lahoratory space and in the open outdoor environment [72, 113-1 17, 137-139, 18YJ. An ohvious criticism of such tests is that even a 1.5-in gap is at least twn orders of magnitude snialler than tlic height of a cloud above grouud from which a typical lightning stroke originates. Such a large

JanuaryiFebruary 2000 -Vol. 16, No. 1 11

extrapolation has been considered by some to be uiiaccept- able and essentially renders laboratory-scale tests useless in evaluating the performance of a lightning rod 1190, 1911.

It can hc argued, liowcver, that for tlie purposes of testing and research 011 lightning rod performance, it is probably not necessary to simulate an entire cloud-to-ground liglitning stroke in the laboratory. It is only required that a realistic simulation he made of the filial leader step in the stroke that approaches a lightning rod 1189, 1911. This reduces the scale and simplifies the problciii enormously, hut still leaves a task that taxes the limitations of present day laboratory facilities. For example, to simulate tlie entire range of striking distances likely to occur in the natural environment, it would be necessary to perform tests using electrode gaps in excess of 100 ni. The gaps presently available in tlie largest lahora- tories are smaller thaii this by roughly a11 order of magnitude. Although some gaiii in gap spacing can be achieved by going to the outdoor etivirotinicnt [l 8x1, one still eiicouii- ters the limitations on voltage imposed by existing impulse generators. Even in the largest laboratories in which ESE devices have beeu tested there is no provision to simulate all of the conditions undcr which lightning occurs in the natural environment. It will be recalled that natural lightning exhibits significant statistical variability in such parameters as current, mean striking distance, and angle of approach with respect to any vertical lightning conductor. It also usually occurs undcr conditions where significant space charge may be present due to local point discharges and where humidity aiid surface moisture levels are relatively high. Moreover, high winds also tend to be associated with the occurrence of lightning. It must he recognized that such parameters as liu- inidity, space charge, and wind are not independent. For example, the rate of space charge development is expected to depend on humidity and tlie wind will he effective in redis- tributing tlie space charge once it is formed [94]. Although tlie influence of factors such as space charge [I921 and humidity [166] have heen iuvestigated iu tlie laboratory tests, it is not clear that the myriad conditions that can exist in the ilatural atmosphere during a thunderstorm can lie ade- quately simulated in present laboratory facilities. The extent to which it may he necessary to simulate all conditions is certainly a snhject for debate.

Another concerti about tlie validity of small-scale simulations is the degree to which tlie discharge produced is like lightning. There is evidence, for example, that the cm-rent aiid propagation velocities of Iahoratory-generated leaders differ considerably from those associated with natural lightning. It would appear that more investigations into compari- sons between the properties of simulated aiid natural discharges may he required before more reliance is placed on laboratory scale testing to evaluate tlie performance of air terminals. The adjustment of laboratory parameters to produce long sparks that match the characteristics of natural lighttiiiig assumes a complete kni~wlcdgc about the characteristics of natural lightning. Tlie extent to which our ktiowl- edge of lightning is sufficiently complete is still open to

debate. The similarities and differences between iiatural lightning and long sparks produced in the laboratory have been extensively discussed (66, 113, 176, 190, 191, 193-2061.

Despite the present limitations, laboratory tests coupled with fast electrical and optical diagnostics probably offer the hest iiieaiis for learning about the physical inechanisms, operation, aiid performance of lightning protection devices in a reasonable time frame. Laboratory tests are especially useful for investigating factors that influence streamer initiation from an air terminal, and thcre seems to he much that cm he learned about the initial discharge growth process for both conveiitioiial and ESE terminals. Many laboratories have been set up to simulate differelit aspects of liglitniiig 192, 207-2171, and even though most of them are not designed specifically for testiiig lightning protection devices, some of tlie advanced diagnostic methods developed in these lahora- tories might find application in air terminal testing.

Because of tlie large statistical variahility in lightning bc- havior, it is unliliely that a single test coufiguratioii can be used to completely characterize the performaiicc of all lightning protection devices. In the future, it will probably be necessary to consider a set of laboratory test configurations that represent die range of lightning behavior likely to be eti- countcrcd in the environment. At present it would seem that we are a long way from having a standard laboratory test procedure for lightning protection systems. Tlie influences of such parameters as moisture, space charge, and wind are still topics for research.

It is recommended that caution he exercised in drawing significant quantitative co~iclusioiis about tlie comparative performauces of different lightning protection systems in the natural environment from small-scale tests. There would especially lie reason to doubt results from siinultaneous tests of two or more devices that arc placed in close enough prox- imity to be within each others supposed range of protection. Under such conditions, the presence of one device can significantly modify the electric field configuration of another device (atid vice-versa) and thereby affect its performance.

SIMULATIONS USING THEORETICAL MODELS With the advent of high-speed computing, it has become

feasible to consider the use o f theoretical simulations of lightning as a tool in evaluating tlie performaiicc of a liglitning protection system. In tlie past 20 years, considerable progress has hccn made iii understanding tlie inechaiiisnis of electrical discharge initiation [17, 67, 169, 170, 218-2221 atid in tlic modeling of corona streamer-leader discharge propagatioiiitibotlisinall and long air gaps [3,59,223-23.31. Nevertheless, the theory of lightning is still in the develop- mental stage and new results continue to appear in the literature. At present, a “statidard” model for the lightning discharge does not exist. The existing iiiodels employ many simplifications and approximations that cannot be exain- iiied or critiqued in this report. It suffices to say that they arc generally designed to accoimt hest for laboratory-scale ob-


scrvationsof long-gapspark dcvclopnieiitiii airaiid havenot reached the level of sophistication required to account for the broad range of complex statistical and fractal behavior characteristic of natural lightning. The stochastic behavior of relatively simple phenomena such as electron-avalaiiclic development and corolla has only recently been dealt with i n thcorctical models [11, 38,2221.

In as much as the effcctivcness of ESF, devices is attribut- able to their ability to enhance initiation of an upward streamer i n the field of an advancing leader, models used to estimate their performance compared with conventional terminals must deal with the statistics of discharge initiation, i.e., they must he capable of predicting time lags applicable to enviroiiniental conditions of the terminal. Unfortunately, the prohlein of statistical time lags is very complex and is generally avoided i n existing computer incidels of discharges. The complexity of the problem is clue in part to a lack of lmowledge ahout microscopic processes of electron release and the statistical behavior of electron-avalanche growth in nonuniform electric fields, particularly under the multitude of conditions that could be encoinitered at the tip of a lightning terminal.

One area whcrc theory shows promise for evaluation of lightning terminals is in the prediction of striking distances 173, 3 86, 193, 234-2461, Assuming that the electric-charge distribution within the approaching leader step is known, cs- timates can be made from electrostatic-field calculations of the instantaneous field at a nearby conductor, e.g., a vertical conducting rod. Assuming a relatively simple charge distrihution in the leader channel, e.g., a linearly nnifurni cylin- drical distribution, it is uften possible to express the field at the teriiiiiiill due to tlie leader in closed form 12401. Calcnla- tions of this type which ta lc into consideration the wide range of possihle leader conditions (defined by such parameters as charge, length, and position) could he useful in esti- mating the maximum ranges of protection 1681. In essence, such calculations, when coupled to the streamer inception criterion 18, 111, 137, 159,223, 225,247,2481, determitic the locations where ai1 advancing leader produces an electric-field strength at the terminal tip sufficient to allow streamer development. It must be understood, however, that sucli calculations supply geometrical information tliat is applicable to all terminals uf the same general geometrical configuration independent of whether or not they are equipped with an ESE device. I n this respect, they do not yield specific information ahout performance of the ESE device itself uii- less the device operates in soch a manner that an impulse voltagc is applied to increase the potential of the tcrniinal conductor tip relative tu ground during the approach of a lightning stroke. Such ii voltage would create a field that adds to thc leader field thereby increasing the presumed maximom range of protection. (It is not clcar that any elec-

d ESE devices actually do this, and i f they do, no information could he found in the literature to indicate that calculations of the type mentioned above have ever been performed.)

Criticisms that cm be raised ahout present striking dis- tancc calculations rclatc to their semi-empirical nature and the fact they asstiiiie leader charge distributions that may he unrcalistic os have not heeii confirmed by observations of actual lightning discharges. Moreover, these calculations have not dealt with effects of atmospheric space charge near ground level.

Validatim of lightning models by comparison of calcu- lated results with observations made clnriiig natural lightning strikes presents a major challenge to theorists. A complete model should account for all observed properties of lightning such as measured current, optical and radio-fre- qucncy emission spectra, propagation velocity, etc. The challenge is made especially difficult by thc broad statistical variahility in lightning behavior and tlie paucity of complete scts of ohservations on single lightning strokes. Neverthe- less, it would appear that tlie method of predicting striking distances hy simulating tlic effect of an approaching leader shows great promise and it should he pursued and improved npon. It possibly offers the hest approach tu answering sonic of the difficult and controversial questions associated with realistic determinations of protection zones that will be discussed later in this report.

TESTS USING NATURAL OR ARTIFICIALLY TRIGGERED LIGHTNING

Perhaps the easiest and least controversial method of testing lightning protection systems is to observe their pcrfor- niancc i n the natural environment during actual thunderstorms. However, this approach is ncither as easy nor as lacking in controversy a s it may first seem. First of all, with the exception of unusually high towers such as the F,n- pire State Building, lightning strikes tu any given location in relatively flat terrain whcrc a lightning rod is positioned are likely to be extremely infrequent 12491. Even in pla experience a high rate of lightning strikes such as in some parts of central Florida, the ninnher of recorded cloud-to-ground strokes within a square kilometer is likely to be less than five per month on average (40, 2.501 during the pcak of the thunderstorm season. Clearly, if natural lightning strikes a terminal only one or twn times per year, it talcs a n extremely long time to acquire enough data on its performance to be statistically meaningful.

Recent attempts to test air terminals positioned at high elevations on iiwutitaiii tops in New Mexico whcrc there is a known high frequency of lightning have shown that lightning seldom hits a terminal regardless of whether or not it is equipped with an ESE dcvice [ 126, 177, 2.511. Although a few isolated strikes to the mountain were reported to have occurred within the supposed zones of protection of ESE ter- ininals 1126, 177J, it would appear that the overwhelniing majority of strikes to the mountain were at considcrahle distance from any terminal. In any case, the failure of air ternii- nals to attract lightning on niountain tups at elevations of 3000 in (9843 feet) or more is obviously disturbing and raises questions about the interpretation of such observa-

JanuaryiFebruary 2000 -~Vol 16, No. 1 13

tions. Before any serious conclusions arc drawn ahout the performance of lightning attractors from tests performed o n inountain tops, it may bc necessary to consider the per- turbing effect of the mountain itself on such parameters as the surface charge distrihution and electric-field profile under a tliundercloud, as well as tlic extent that lightning strokes at such high elevations differ from those that nor- mally occur in lower, flatter locations. It would appear tliat the aiiswcrs to some of these questions might already be found in the literature.

It is noted in sonic papers that lightning that occurs at high elevations generally differs on average from that wliicli occurs at sea level, if in no other respect than that it lias less distance to cover in going from the cloud to the ground L271. At an elevation of 3000 m, tlie ground can he quite close to or even engulfed hy tlic base of a storm cloud. Certainly tlic results from high mountain tests cannot be dismissed, aiid such tests should continuc, a s should similar tests underway at other locations 11 761. The problem is how to iiiterpret the results of these tests and infer what they might imply about air terminal perforinatice at lower elevations, and what they indicate about the influence of nionntaiiious or rocky terrain on the effective zone of protection of a n air terminal.

The inifavorablc statistical odds associated with natural lightning can be partially overcome hy using artificially triggered lightning. Tests have shown that lightning can be triggered with reasonably high probability by a rockct Iaunclied into a thundercloud 185, 2.52-2.541. A long trailing wire is usually attached to the rocket, which provides a low rcsis- tatice path to guide tlie initial discharge and define its direc- tion of propagation [U, 255, 2561. Transportahle facilities have been dcvclopcd for rocket triggering of lightning that can he used for testing at nearly any location [257]. Although tests of air temiinals are bcirig rnadc using trigerred lielit- ning, there arc questions that can he raised about the meaii- ing of such tests. There is evidence that triggered lightning is unlike natural lightning both in its intensity and propagation characteristics. In particular, it has heen noted that triggered lightning is of Iuwcr current than natural lightning and exhibits characteristics inore like those of return strokes oh- served i n natural lightning [25R, 2.591. It lias also hccn argued that triggered lightning does not satisfactorily mimic tlie primary stroke and is thcrcfore unsuited for investigation of the nttachmcnt to 3 grounded lightning conductors, i.e., its use i n evaluating air teriiiinals would appcar to be questionable [258]. The extent to which rocket-triggered lightning behaves like natural lightning seems to depciid oii the length of the trailing wire and the distance of the hottum ctid of the wire ahove ground when the discharge occurs. Notwithstanding valid criticisnis, essentially no quantitative information could he fouiid in the literature about results from tests performed on air terminals using artificially triggered lightning I26Ol.

Even thoogh testing of air terminals using natural lightning lias obvious limitations, we would rccoiiimcnd long-term or continuous monitoring of lightning around air

terminals during thurtdcrstorru activity. Data from such monitoring could provc valnablc i n identifying conditions under which lightning protection devices arc likely to fail. Admittedly it is difficult to draw meaningful conclusions froni isolated events. However, it can be argued that previ- ous liglitning records [36,43,5 1, 1761 have proven useful in revealing unusual forms of lightning behavior that ought to be considered in designing lahoratory methods o r computer simulations fur use in evaluating lightning rods.

ISSUES TO BE ADDRESSED In concluding this discussion we draw attention to four

specific issues: ‘ I ) tlic zone o f protection, 2) uncontrollable factors, 3) radiation hazards, aiid 4) damage and maintenance. For purposes of evaluating the relative performance of ESE and conventional air terminals, the zone of protection is by far the most important. The second issue is connected with the first in the sense that thcrc may exist uocontrollablc factors that affect the zone of protection offered by a terminal regardless of whether or not it is cquipped with an ESE device.

Zone of Protection The classical “coiie-of-protectioii” concept is often used to

specify tlie regiou of space that is “protected” by a lightning conductor. This coiiccpt was first iutroduccd i n the nine- teenth century and is based on rather simplistic electrostatic field analysis using a rod-plane type geometry in which the base o f the tliuudcrcloud is assumed to have a uniform charge distrihution [15, 2611. More recently the “rolling sphere” method has heen introduced [14, 16, 122, 178, 2.5811 to estimate protection zones. Although this inethod call be vicwcd as an cxtcnsion uf the c[)nc-[,f-protectioii concept, it goes beyond this concept in providing identification of possible attachment points within tlie cone. The rolling sphere method allows for possihlc lightning strikes to the side of tlie Knipire State Building, whereas the cone-of-protection does not.

Despite their simplicity, these concepts can he used to male first order estimates of protection zones arouiid a lightning rod or an array of lightning rods [ U , 122, 178, 262-2641. However, we would judge that zone-of-protcction cstiiiiatcs tliat arc derived from electrostatic field calculations are easily misinterpreted and can lead to exaggerated or unreaJistic claims about the protection capalilitics of air terminals (regardless of whether or not they are equipped with a n ESE device).

Recognizing tliat lightning is a stochastic proccss that exhibits a broad range of hchavior, it has been recoininended that the simplistic zone-of-protection concept be replaced with a more realistic statistical dcscriptiun in which, for example, the most prohahle (or maximum) striking distance is displayed graphically as a function of leader or primary stroke currcnt [22, 239, 265, 2661. It has been noted that striking distances for positive discharges will differ from those for negative discharges, and that tlie knowledge about positive striking distances is inadequate [13].


Nevertheless, enough may uow be known about the hchavior of lightning and the lightning attachment proccss that inore sophisticated and statistically meaningful statements can be made about protection zones than are used presently [68, 193, 258, 2671.

Recently developed models for calculating striking distances [73,111, 186,235,237,239-246,2681 could prove useful, for example, i n determining tlie niaxinium distances from a terminal at which leaders with particular cliaracteris- tics (length, charge distribution, and velocity) could enhance the local electric-field strength enough to allow dcvclopmcnt of an upward streamer. Such calculations place an up- per bound on the size of the protection zone for a terminal with a given geometry (height) indcpendent of its ESE cliaractcristics 12691. Given knowledge (or assumptions) about enhancement of streamer initiation probability at an ESE terminal for a particular local field strength, it is conceivable that reasonable quantitative estimates could bc made of the incremental increase in liglitning attraction efficiency of ESE terminals over conventional terminals. No evidence could be found that this type of analysis has ever been attempted for ESE terminals. Most of tlie theoretical work 011 striking distance has been motivated hy the concerns of clcctric-power utilities about lightning strikes to power trans- mission systems [14,23,236,239-244,2691. Much of what has been learned from tlie utility work can undoubtedly be applied to an evaluation of air terminals.

Uncontrollable Factors One of tlie most difficult prohlerns faced in the design of

air terminals is associated with assessing the influence of 1111- controllal~le factors. Included here are the effects of nearby objects such as trees, builtlings, smoke stacks, etc. Such objects may not only he sources of corona and therefore spacc charge [94, 2701; they may also significantly perturb the electric field within the specified zone-of-protection. In ad- ditiun to nearby objects, the terrain itself can also be a factor in deterinining realistic zones of protection. If the effective zone of protection is extended througli the use of an ESE dcvice, then problems of assessing the influence of other objects and variations in terrain arc also extended.

Flying debris in the vicinity of an air terminal (dust, leaves, sticks, paper, etc.) may also be of concern, particularly if it can somehow attach to the terminal. This conccrn is justified because high winds often associated wit11 tliunder- storms stir up and elevate ground matter and because the more complex geometries used in the construction of ESE devices may offer greater opportunities for trapping ground matter. The effect of flying dchris is an issue that s e e m to be ignored in tlie literature. On the other hand, no evidence could be found to suggest that this effect is responsible for any failures of air terminals to attract lighttiing.

Radiation Hazards In the case of ESE devices cnipluying radioactive materi-

als, issues have bccn raised about tlie possible radiation Iiaz-

ards to humans that the use of thcsc devices present [I 12, 129-132, 135, 2711. As noted above, radioactive air ternii- nals are banned i n some couutries, presumably because of perceived health hazards. It has been noted that 241Am sources used in lightning protection devices are not any more hazardous than similar sources approved for use in smoke detectors or static eliminators [104, 112, 1341. Ncv- ertheless, there are those who argue that the public may be placed at risk from a proliferation of radioactive materials in devices that can enter the enviromnent without adequate contruls [112, 131, 2711.

Damage and Maintenance Given that ESE devices likely have a structure and associ-

ated instrunientation that arc niorc complex than conventional air terminals, questions can be raised about tlicir susceptibility to damage during a lightning strike. The electric current and energy deposited by a lightning stroke can be sufficiently high to actually melt metallic structnres and de- stroy electronic components. Tliere arc numerous reports of damage inflicted hy tlie primary lightning stroke to metal parts on aircraft, etc. [211, 272-2761, The possibility of damage means that a lightning protection device may re- quire periodic inspection and/or niaintenance that is gcnerally not required for conventional terminals. Although this problem is pointed out [105], there seems to he very little discussion about it in the open literature.

Conclusions The possible coi~clusioiis that can be drawn from an ex-

amination of tlie literature are discussed in this section. The main conclusions of this report are briefly summarized in tlie Summary of Conclusions.

Due to the paucity of reliable quantitative data from tests of ESE air termiiials that can be found i n tlie peer-reviewed literature, it is nearly impossible to make quantitatively mcaningful statements on the relative pcrforinance uf ESE devices and conventional Franklin rods. In fact, sufficient rc- liable quantitative data on the pcrformancc of conventional rods seen- not to exist. There is, for conventional terminals, ongoing debate about the best geometry for optimal lightning attraction efficiency, for example.

Nearly all of the data found on ESE device performance resulted either from tests performed by manufacturers of lightning protection systems or by those directly or indi- rectly employed by such manufacturers. Altliougli criticism is published by non-manufacturers about tlie performance of ESE devices, especially radioactive air terminals, it is seldom based on actual test data. Those on both sides of the issue iti- voke lack of evidence in making their case abont the performance of ESE terminals. I’roponents of these devices claim that a lack of credible statistical data on failure of ESE terminals proves their effectiveness, while critics of these tcrmi- ilals argue that a lack of evidence about t l ic improved perforiiiance of ESE terminals over conventional tcrtninals proves their incffcctivcn In citlicr case, one must beware

JanuaryiFebruary 2000 -Vol. 16, No, 1 15

of faulry logic, in as much as a lack of cvidciice never proves t l i e lack (if something.

There arc reports of incidents where ESE devices failed to provide the protectioii specified b y tlie manufacturer [ I h , 126, 277, 27x1. Statistics on tlie failure of conveutional sys- t e m have also hccn documented [104]. Wheu examining reports of “failures,” one can always raise questious about their cause, e.g., whether they are primarily a conseqociice of exaggerated claims made by tlic manufacturer or a couse- quencc of misuse (faulty installation) of tlie device. Reports of isolated failures raise legitimate couccrus, hut are seldom accoiiipanicd h y eiiough supporting data about the event to enable a dctcrmination of why tlie failure occurrcd. Gen- erally it is difficult tu draw significant conclusions from single events that cati he used to improve system desigu or evaluate system perforniancc. There is no reason to believe that an air terminal is 100% efficient in attracting lightiiing, regardless of what kind of ESE device it uses, i f auy. Con- sidering the wide range of possible atmospheric conditions and types of liglittiing behavior that have been recorded, it is uot surprising tliat air teriniiials of all types will sornetimcs fail [46,49,53 1. Tall structiircs arc reported to be struck oc- casionally by lightning at points far below die top, i.e., outside of the “protcction zone” 145, 52, 2791. Any claims o l 100%1 efficieucy i n the performatice of a lightning attractor should he viewed with sliepticism. In any case, tlie uicaning of tlic term “efficiency,” wlieti specified for a n air terminal, should be clearly defined and understood.

A rcasotiable physical basis fur tlie operation of a11 ESE device appears to exist in the sense that there is good evidence from laboratory investigatioiis that the ps(hahi1ity of iuitiatiug a streamer discharge from a11 electrode caii be ill-

creased significantly by irradiation or electrical triggering. However, tlie precise amomit by which this enhancement in streamer initiation improves tlie lightning attraction efficiency of an air terminal remains questionable. Thcrc is rea- sou to doubt tliat it significantly extends the maximum range of protcctiou. A lightning stroke that would not liit a conventional terminal because of the fact tliat it does not en- hancc the field at the terminal tip enough to allow streamer iormatioti will also iiot likely liit a termiiial equipped with a11 ESE device. (The exception would be an ESE device tliat significantly iucreases tlie terminal potential during the approach of a lightning stroke.) In our view, the possible advantage offered by an ESE device, if operated properly, is that it helps to iusure that a streamer will be initiated if tlie field produced hy tlie oucomiiig stroke a t tlie terniiiial bc- conies sufficietitly great to allow streamer propagation. When comparing ESE aiid cotiveiitioiial terminals, it is probably preferable to consider efficiency rather than zoiic of protectioii as a measure of pcrforinance.

It is possible that tlie increase in liglitiiing attraction efficiency gained from usiug an ESE device can also be achieved by simply using a cotiventional termiual of greater height [113j. There is no iudisputable evidence from the literature that die range of protectioii offered h y a single ESE terminal

cati be greater than or necessarily tlie same as tlie range pso- vided by two o r more conventional terminals of the same height with overlapping zoues of protection. On the other hand, an array of ESE terminals may provide better protectiou than a similar array of coiiveiitional devices. Altliougli tlie precise amoutit by which the ESE device extends or iiii-

proves the performance of a conveiitioiial terminal is geuer- ally not known or easily measured; there is no reason to believe tliat an ESE array will have an iuferior performance. We would argue that until issues concerning the relative performances of siugle ESE and coiivcntional terniiiials are set- tled incaningful statements caunot bc made about the coinpnrative performances of arrays of these terminals.

111 geiieral, it can bc presumed that ESE terminals perform at least as well as conventional tcrmiuals with tlie same geometrical configuration provided, of course, that they are properly designed to avoid significant coroua formation during a thunderstorm. In tlie event that an ESE device fails or hccoines inoperative for some reason, the ESE terminal sliould revert in its characteristics aiid performance to that of a conventional tcruiiual with tlic same height, geometrical configuration, and conuectiou to ground.

Much has been learned about the operation of ESE tcrmi- iials from laboratory-scale tests to suggest that ESE devices do indeed etihaiice streamer cmissiou compared to convcntional terminals. However, these results have uot been, and probably cannot be, used to make quantitative dctcrmina- tions of the relative cfficieiicies of these terminals for atmospheric conditions utider a thundcrstorm. At tlie present time, tlie results from a limited tiimiber of field tests with uatural lightning are inconclusive with respect to providing estimates of relative efficiencies. It is iiot clear that enough data call ever be acquired from such tests to draw quaiitita- tivc couclusions about attraction efficieucy. Tests in the iiatural environment appear to be most useful in identifyiiig and documenting conditions uuder which ais terminals fail.

Semi-cmpirical models have reccntly hcen developed to calculate striking distances to lightning conductors. These models show promise in providiug a method for making realistic estimates of maxiruum protectiou range for air terminals. ‘The maximum extent to which ESE devices enhatice tlie attsactiou efficiency or increase tlie effective range of protection of a termitial could conceivably bc iuvestigated with these models.

Recommendations Until recently, because there is not much that can be done

to improve tlie design of a conventional liglitning rod, tlicrc was little motivation to perform complicated tests to evalo- ate tlic efficiency of these rods as liglimiiig attractors. It has always been rccoguized that conventional rods soiiietinics failed and that the rcasons for failure were usually attrihut- able to tlic complex atid unpredictable uature of tlie lightning discharge. With the appearatice oti the market of competing products (ESE devices that supposedly improve the attractim cfficieticy of a rod) have come questions about


how these new devices work and how they can he tested to verify their performance. Lightning, in the meantime, remains as complicated as ever, and our understanding of lightning seems to progress only very slowly. Given this situ- ation, it is not clear that we can find quantitatively accept- able answers to questions about the performance of lightning protection systems any time soon.

Considering the difficulty of the task, we offer the following recommendations for future work without being specific about realistic timetables and expectations:

1. Give priority to developing 11cw methods for calculating or otherwise determining striking distances and related zones of protectioii that are more meauingful from a statistical point of view.

2. Continue and extend laboratory tests to investigate the effects of relevant parainctcrs such as polarity, space charge, wind, and humidity OD the streamer initiation probabilities and propagation from ESE and coiiventioual terminals.

3 . Continue observations of natural lightning in and around test sites setup with different air terminals in various locations where the frequency of lightning is knowu to be high.

4. Compile atid analyze existing and newly acquired statistical data on tlie behavior of lightning from different locations and different sources in a central location. T o enhance credibility, more of the testing and data evaluation should be pcrforined by individuals or organizations not identified with manufacturers of lightning protection systems.

Summary of Conclusions The main conclusions of this report can he summarized in

tlie following statements: 1. Lightning is a complex, chaotic phenomenon that ex-

hibits a broad range of behavior and characteristics that arc unpredictable. Any theory or assessment of lightniug protcction devices must take this fact into account.

2. A plausible physical basis exists for ESE devices, in the scnsc that they can enhance the probability for initiating an upward propagating streamer, which is directed at an 011- coining lightning stroke, from an air terminal.

3. Insufficient iiiformation could be found ahout both ESE and conventional air terminals to allow a ineaningful comparison of their relative performance i n a natnral clivi- ronment.

In conclusion, it could be said that there is yet more to bc learned about lightning and about how lightniug protection devices work or do not work. The road to better lightning protection is obviously strewn with controversy and it would appear that the path to resolution requires more en- lightenment and less thunder.

Acknowledgments The study was completed with support from the Technol-

ogy Administration of the U.S. Department of Commerce and was funded in part by the National Fire Protection Ke-

search Foundation. The authors thank Samara Firebaugh for her contributions to the work.

Richard J. Van Brunt received his B.Sc. in physics from the University of Florida in 1961 and 1’h.D. in physics from the University of Colorado-Toint Institute for Lahoratory As-

Group of the Elcctricity Division. He is an IEEE fellow and was the IEEE Whitehead Memorial Lecturer i i i 1994. He has coli- ducted cxtensive research in atomic and molecular physics, with an emphasis on gaseous dielectrics, plasma processing, and partial dischargc detection.

Tom L. Nclson received a B.Sc. in electrical engineering from North Dakota State Univer- sity in 1988. In 1988 he joined the Electrical System Group of the National Institute of Standards and Technology, Gaithersburg, MD, where he has been working in tlie areas of precision power and energy measurements.

Kenneth L. Stricklett rcccived his B.Sc. in mathematics from tlie University of Ne- braska-Lincoln in 1979, and a M.Sc. and P1i.D. in physics from the University of Ne- braska-Lincoln in 1981 and 1987, respec- tively. He then was a NORCUS post-doctoral Fellow at the Pacific Northwest Laboratories. He has been a researcher at the National Insti-

tute uf Standards and Technology in the Electrical Systems Group of the Electricity Division since 1989. He is an IEEE member. He has conducted research in atomic and niolccular physics and in electrical hreakdown and partial discharge dc- tection in fluids.

References

.I : y .

1. 11.J. Van Brunt, T.S.. Nelson, and S.L. I’irebaugh, ‘‘Early strcnincr cinission lighming prorcctioii sysrcms-literature survey and tccliiiicill evaluation,” NlSTlR 562 I , National Institute of Standards and l’echaolugy, Jan. 31 1995.

2. “I.ighming reference hihliugtaphy 1936- 1949,” American Institute of Electrical Engineers, 33 Wcst 39th Stveet, New York, NY, Apr. 19SO.

3 . M. Uman and E.P. Kridcr, “A veview of iiatural lightning: Ekperimcutal data and modeling,” IFEE Trans. Bkctromag, Comp., Vol. EMC-24, pp. 79-112, May 1982.

4. M. 1. Kalinin, ”Mechanism uf curona-to-spark triinsitim iii long air gnps,”Sov. Phys.-Tech. l’bys., Vol. IO, pp. 948-95 I, J a n 1966.

5. I<. Klingbeil and D.A. Tidmnn, “Theory and C O I I I ~ U ~ D T iiiu&l of tlw lightning stepped Icadcr,”]. Gco. Res., Vol. 79, pp. 865-X69, 1:cb. 20 1974.

6. L.B. S.ocb, “Conh in r ion and cxtcnsion of n proposed mcchanisiu of the stepped leader lightning strokc,”]. Geo. Res., Vol. 73, pp. 58 13-5817, Sept. IS 1978.

7. 1 . Gallimberti, “Acoqmter model for st iemier prupngatii,n,”J. Phys. I): Appl. Phyi., Vol. 5, pp. 2179-2189, 1972.

8. J.M. Mcrk, “A rhcory of spark dischargc,” Phys. Rev., Vol. 57, pp. 722-728, A ~ ~ . i s 1940.


9. Y.T. Lin, M.A. Unum, and R.B. Standler, “1,ightning vetum stroke

10. R. Thottappillil and M.A. Umnn, “1.ightning I’ctwn stroke model with height-variable discharge time constant,” /. Geo. Res., Vol. 99, pp. 22773-22780, Nov. 20 1994.

1 1. 11.J. Van Bmnt, “Physics md chemistry of partial discharge and corona-Recent advances and future challcngcs,” I Electr. Insul., Vol. 1, no. 5, pp. 761-784, 1994.

12. R.H. Golde, Lightning Conductor, Vol. 2. Acadeinic Press, New York, 1977.

13. I1.H. Golde, “Lightning and tall structumS,” 11iE I’roc., Vol. 125, pp.

347-351, Apr. 1978. 14.1<.H. LCC, “Protcctian zone for buildings against lightning stmkcs using

trnnmiission linc protcction practice,” IEEE Trans. lid. App., Vol. IA-14, pp. 465-470, Nov. /Dec. 1978.

15. W.H. Prercr, “On the space pratcctcd by a light~~ing-clind~Icror,” Phi. Mag. J. Sci., Vol. 10, pp. 427-430, July-Dec. 1880.

16. D. Mackerras, M. Darveniza, and A.C. I iew, ”Standad and now standard lightning protection methods,”/. Elcc. Electr. Rng. Australia, Vol. 7, pp. 133-140, Junc 1987.

17. M. Khalcrl, “Computation of COIOIEI onset using thc ring-chnrgc method,” IEE Proc., Vol. 122, pp. 107-11 1, Jan. 1975.

18. M. Meiies and L.H. F i s h , “l’ositive point-to-plane COIUIII studies in air,”Phys. Rev., Vol. 94, pp. 1-6, Apv. I , 1954.

19. C.G. Morgan, “Irradiation and time lags,” in Electrical Urmkdotun of

Gases (J.M. Mcck andJ.D. Cmggs, eds.), pp. 655-688, Ncw York, N Y John Wiley & Sons, Ltd., 1978.

20. R.B. Carpenter, Jr., “Lightning prcvcnti~,n-Practicnl and pSOVe11,”

Measuremorts atrd Data, Vol. 10, pp. 90-96, Jan. /Feb. 1976. 21. J. Hughes, “Review of lightning prorcctiuii technology for tall struct-

ures,’’ tech. rep., Office of Naval Ilesearch, Arlington, Virginia, Jan. I , 1977.

22. R.H. Goldc, “The lightning cmductor,”/. Frati. Inst., Vol. 283, pp. 451-477,June 1967.

23. M.A. Coodlet, “Lightning,”/. Inst. Elect. E,rg.,Vol. 81, pp. 1-2l,July 2937.

24. M A . Uman, “Natural lightning,”IEEB Trans. h d . App., Val. 30, trio. 3, pp. 785-790, 1994.

25. M.A. Uman, The Lightning Discharge, Vol. 39. Academic Press, Inc., New York, NY, 1987.

26. C.F. Wagner and G.D. McCann, “Lightning phenomena,” Ulec. h g . , pp. 1-35, Aug.-Oct. 1941.

27. K. Bergcr, “Measurcmcnts and results of the lightning invrstigatim of

the years 1955-1963 on Mount San Snlvatare,” Ilull. SwissUlektrr,tcch. Ver., Vol. 56, no. 1, pp. 1-23, 1965.

28. C.E.R. llruce and K.H. Golde, “The lightning discharge,"/. Inst. I<lec. E q . 11,Vol. 88, pp. 487-520, Dec. 1941.

29. V. Cooray and 61. Perez, “Some featiircs of lightning flashes observed i n

30. J.H. Hagenguth, “Photographic study of lightning,” Traits. Ain. Inst. Elec. Eng., Vol. 66, pp. 577-585, 1947.

31. J.H. Hagcnguth and J.C. Anderson, “Lighming to the Empirc Stntc Building Part 111,” Trans. Am. brst. E I a . Enfi., Vol. 71, Aug. 19.52.

32. J. Hojo, M. Ishii, T. Kawamura, F. SuLoki, H. Komoro, and H. Shiogama, "Seasonal variation oi cloud~to~gruond lightning flnsh characteristics in thc coastal area of the Sea ofJapan,”]. Gcu. Res., Vol. 94, pp. 13207-13212, Sept. 30 1989.

33. V. Mazur, “Lightning channel properties dctcnnincd with n ucrtic;dly pointing doppler radar,” J. CLO. Res., Vol. 90, pp. 6165-6 174, Juoc 30 1985.

mociris,=j. G ~ ~ . K ~ S , vol. 85, pp, 1571~1583, ~ ~ r . 20 1980.

swrcien,~’j. G ~ ” . K ~ ~ . , vol. 99, I’p. i0683-1068~, M~~ 20 1994.

34. K.B. McEachron, ”lightning to tlic F,mpirc Statc Iluilding,” ‘ l h n s . Am.

35. R.F. Orvillc and V.P. Idone, “Lightning leader clialacteiistics i n die thutidcrstonn rcsenich international program (TRIP),” /, Geo. Iles.,

36. H.M. ‘ I ’ lmns~n , “Characteristics of Port Morrsby gmond flashes,” /.

37. G,W, Trichel, “The mechanism uf tlic point-tu-planc C U I U ~ iii air at

38. 1I.J. Van Brunt and S.V. Kulkami, “Stochastic propcnics of Trichel- pulse corona: A noti-Mhlnrkovian mndoni point proccss,” I’hys. Rev. A,

39. K.N. Bard and D. ~Mackerras, ‘‘l’usitiue cloud~to~groond lightning discharges i n Kathmaitdu thondel-srorms,” /. (;eo. Res., Vol. 28, pp.

103.3 1.10340, Jone 20 1993. 40. K.F. Orvillc, “Cl~,ud-t~,-gn,oi,‘I lightning f lash charactwistics iii tlic

cuntigoous Uoitcd Statcs: 1989-1991,” /, Ceo. Res., Vol. 99, pp. JOX33~10841, May 20 1994.

41. W. Bcaslcy, ”Positive cloud-to-grouud lightning obsevvations,”]. Gea.

42. C.G. Kittcrmnn, ”Cliaractcristics of lightning from frontal systeni thonilcr\tornls,” eo. R~s . , VOI. 8 5 , pp. 5503-5505, 0ct.20 19x0.

43. S. Szpor, “Comparison of I’olish versus Ainericnn liglitning ~.ecards,” IEEE Tuairs. Power App. Sys., Vol. PAS-88, pp. 646-652, May 1969.

44. W. Janischewvskgi, A.M. Hossein, 1’. Dziurewicz, and V. Slmatak, “CI~~,l.nctel.ization of the C L I I T C I I ~ wavefront parameters of lightning str ikes to the CN Towel- i n Toronto,” i n 8th Iirtenmtiurial Symposium on l~ligh Voltage Enginccrifil’I.ocecdrnfis, Vol. 3 (Yakahama,Jnpnn), pp. 221-224, IKP. of]apan, Ang. 23 1993.

4.5. I<,ll, McEacliron, “Lightning to the Empire Srate Building,” J. Frati. Inst., Vol. 227, pp, 149-217, Pch. 1939.

46. K. Rerger, “Nwcl ohscrvations 011 liglitiiing discharges: Results of rraral-cl~~inMr~untSanSalvatorc,”J.Fran.lnst.,Vol. 283,pp.478-525, June 1967.

47. M.A. Snrgcnt, ”Thc frcqucncy distdmtion of CLirrciit (magnitudes of

lightning strokcs to tall structiircs,” 11~a>~s. I’oioer App. Ys., Vol. PAS-91, pp, 2224~2229, Sept. /Oct. 1972.

48. B.F.J. Schanland, “lkvelopment of tlir lightning diachnrgr,” Nut., Vol.

49. A. Uman, W.H. lleasley, J.h. ‘Tiller, Y. Lin, B.P. Kridel-, C.D. Wcid- mai~n, P.ll. Krchbicl, M. llraok, A.A.J. Few, J.1,. Ilohannon, C.L. Lcnnun, H.A. Pochlcr, W. Jnffcris, J.R. Chilick, and ].I<. Nicholson, “An uiiiis~ial lightning flash at Kcnncdy Spacc Ccntcr,” Sci., Vol. 201, pp.

Inst. ~ 1 ~ ~ . E S ~ . , vUi. 60, pp 885-889, scpt. 1941.

vol. 87,

c ; ~ ~ . K ~ ~ . , vol. 8.5, I,l,, 1027.1036, pel,. 20 19x0.

I I 177.1 I 192, DCC. 20 1982.

atllluSphC~K prCSSWC,”I’hyS. Re”., Vol. 5 5 , pp. 382-399, 1939.

vol. 42, p p 4908-4932, Oct. 15 IWO.

iles., v0i. 90, 6131.6138, J,,I.. 30 1985.

1.32, pp. 407.408, Sept. 9 1933.

9-16, J u l y 7 1978. 50. T.L. Teer and A.A. I’ew. “1 lorirunral liehtnina.” 1. Geu. Res., Vol. 79. .. ”_ ,

pp. 34.36-3441, i\ug. 20 1974. 51. R.P. Kridcr, “An i i i i i is~al photogrnpli of i i t i iiit lightning disclwge,”

Wea., Vol. 29, pp. 24-27, 1974. 12, A.M. Muosa, “Effcct of hcight of stmctwc on tlic striking distance o i a

downward lightning flash,” i n Proceedings of bitemational Conr,nu,iications & Enerm Conference (Montreal, Quehcc), pp. 9-1 4, IEEE Publication No. CH20412, Oct. 1 1994.

5 3 . 1I.E. Orvi l lc and K. Ilcrger, “hn ~ I ~ U S L I P I lightning flash iniriated by a n upw.,rd-i”-”l’agatilig leader,” J. Gco. Iles., Vol. 78, pp. 4520-4525, July 20 1973.

54. K . W . Williams, ‘‘TIE rlccrrificatioii of thondcrstorms,” Sci. Am., pp.

5 5 . M.F. Borisov, E A . Zobov, and 1.G. Litvioova, “Lateral nun-unifurmity iii rhc glow oi a Ihigli-power spark channel,” Elektrichestva, pp. 66-68, Oct. 1991.

xn-99, NOV. 1988.


( 6 . V. Luncv, “Luminous balls in Sihcria a i d the Far Fast: I’l~cn~~mctlology, cxpcrimcnt, liy~~~’thrsis,”Sov.l’hys.].,Vol.35,pp.256-274,May 1992.

tekolnikov, “Study of liglitning and lightning protection,” tech.

vep., I’oreign ‘Technology Division, Wright-Pnttcrson AFD, Feb. 18 I9XX.

S X . K. Yasui, “Plnsmn fircballs fed hy ~iiicrowrlvcs,” Phys. Lctt. A, Vol. 173, pp. 45 1-455, Pel). 22 1993.

59. A. Bondiou and 1. Gallimberti, “Theorcticnl modelling of the dcvclop- ment of the posirive spark i n lung gaps,”], I’hys. D:AbpI I’hyr., Vol. 27, pp. 1252-1266, 1994.

60. M.M. Kekez and P. Snvic, ”I.abaratorysimuIatii,n of thc StCpped leader

in lightning, ’’ Can. J. Phys., Vol. 54, pp. 2216-2224, Nov. 15 1976. 6 1 . 5 R. Khastgir, “Lcadcrstrakecurrent innlightningdischargeaccording

to thc strcaiiwr tlieory,” Phys. Rev., Vol. 106, p p 616-617, M t y 15 1957.

62. C.T. Phclps, “l’ositive streamer systcm intcnsificstion and its possiblc role in lightning itiitiation,”]. Atin. ‘1i.r. Phys., Vol. .36, p p 103-111, 1974.

63. B.F.J. Schonland, U.J. Maln, and H. Collens, “l’rugrcssivc lighttiing-11,” Puoc. Roy. Soc. Lotid. A, Val. 152, pi’. 59.5-625, 1935.

64. S. Srpur, “Relaxation tllcory of the lightning StCppCd Icader,” I j Id / .

Assoc. SuisscElcctr., Val. 68, pp. 1293-1296, Dcc. 1977. 65. S. Szpor, “Rcview of the theories of thc liglitoing iiiaiii dischnrgc,”

Axhiwum Hick-trotechniki, Vol. 26, W . 2, pp. 279-290, 1977. 6 6 . C.F. Wagner and A.R. I-lileman, “Thc liglitning stroke-11,’’ Trans. Am.

lmt. Elcc. Ens., 111, Vol. SO, p p 622-642, Oct. 1961. 67. S. Yokoyania, K. Miyake, T. Suzuki, and S. Kanso, “Wintcr lightniiig on

Japan Sea const: Dcrelopmcnr of orcnsuring system 011 progressing featorc of lightning discharge,” 1EEE Trans. Power I)eliuc?y, Vol. 5 , pp. 1418-1425, July 1990.

68. C.F. Wagner and A.R. Hilcmnn, “‘l’he lightning srrokc,” ‘li-nnr. Am. Inst. E l m . ERR., 111, Vol. 77, pp. 229-242, Junc 1958.

69. T.F.. Allihnne and J.M. Mcck, "The development of tllc spark din- chargc,”Proc.Roy.Soc. I.u,id.A,Vol. l66 ,pp. 97-126,June 16 1938.

70. T.E. Allibone and J.M. Meek, “‘lhc ~lcvclopment of tlir spark discharge-11,” I+m. I<oy Soc. L m d . A, Vol. 169, pp. 246-268, Mar. 7 1939.

71. K.H. Galdr, “Occorrencc afupwardstreninrl-s i n Iighttringdischargrs,”

72. 7. Shindo, Y. Ailiara, ;md 7‘. Suzuki, “Mudcl cxpcvimcnl of u p w d leaders-Shielding effccts of tall obiccts,” I Trans. Power Ileliuery, Vol. 5, pp. 716-723, Apr. 1990.

73. M. Abdcl-Salam, M.T. El-Mohnndes, C. Bcrger, and B. Senouci, “Onset criterion of u p ~ ~ d streamers from a Franklin rod,” J. Elcctroslalrcs, Vol. 24, no. 1 , pp. 45-59, 1989.

74. I.. Nicineyer, L. I’ictroncru, and H.J. Wiesmnnn, “Fractal dimcnsioii of

diclcctric breakdi,wn,”Phys. Rcu. I.eif., Vol. .SI, pp. 1033-1036, Mar. I 9 19x4.

75. 1. Arima, T. Wat;,niibe, N. ‘likagi, atid M. Kakilinm, “Exprrimcntal study of thc corona cliratli ciirrcnt in lightning rrtiirii stroke,” iii IX Intcnintionul Confuence on Gas Discharge and their Applicirtioiis

76. I.J. Cnylor, V. Chandrascknr, V.N. Rrhgi , and S.S. Miogcr, “Multi- paramcrcr radar ohsewations of lightning,” in liiternatioiral Confere,rce oit Radar Meteorology ( N ~ r r n m , OK USA), pp. 306-308, Amcricnn Meteordogicnl Socicty, ‘1993.

77. A.S. Gayvaratisky and K.V. Karasyiik, “Nunierical modc l of liglitning lender uricntntim on a tl.aiisinissim liiic,” iii 8th lntematianal ,Symposium on High Voltage Eriginccriiifi Proceedings, Vol. 3 (Ybkohama, Japan), pp. 277-2S0, IEli of Japan, t\ug. 23 1993.

“zt , , Vol. 160, pp. 395-396, SCpt. 20 1947.

(Venice, Italy), pp. 431-434, Sept. 19 1988.

78. b’, Ileidler, “Lcrnp cakulatim and lightning cwrcnt fonction,” in 8th lntemntional Symposium on HiRh Voltqe Lhgimcri,i,q l’roceediiigs, Vol. 4 (Yakohnina, Japan), pp, 237-240, IEE of Japsn, Aug. 23 199.3.

79. R.1:. Criffiths and C.T. P I M ~ s , “A iiiodcl for lightning initiarioii arising from positive C O I . O ~ strcainer deuclopmcnt,”]. GEO. Kcr., Vol. 81, pp.

80. B. Vonnegot, “l‘lw atnmsphriic clectl-icity paradigm,” Uti11 Am. Met. Sac., Vol. 75, pp. 53-60, Jan. 1994.

8 1 , P.R. I<rchhicl, M. Brook, and K.A. McCrory, “hn analysis of t l i c cli:irge structure of lightning discharges to gronnd,”]. Geo. l<es., Vol. 84, pp. 2432-2456, May 20 1979.

82. J . A Cldmem, “l’i,inrdischnigc CIIII.CII~S,”], A m . Ter. I’hyr., Vol. 2, pp. 301-305, 1951.

83. C.B. Moore, “Improved configur.itims of lightning rods and air tmniiials,”]. Fvarr. litst., Vol. 315, pp. 61-85,Jnn. 1983.

84. A.A. Barncs, ”Prcdicting triggercd lightning,” i n Intenmtional Con- ferciice on 1,ightaingnndStnticBIcctricity (Bath, U.K.), Scpt. 2X 1989.

XS. M.M. Newman,J.R. Srahmano, and J.D. Robb, ”Expcrimcotnl s t d y of triggered ~ i a t ~ i l a l liglitning dischal-gcs,” t d i . rep., Fcdcral Aviation Administl-arion, I’rujcct Nu. 520-002-03X, 1967.

86. E.R. Williams, C.M. Cookc, and K.A. Wright, ‘”l‘hc rolc of clcctric spacrchargein noclcarlight~~ing,”]. Geo. I<Es.,VOI. 93,pp. 1679-1688, Fcb. 20 19XX.

X7. E.K. Williams, C.M. Couhc, and K.A. Wright, ”Elcctrical dischargc propagation in and al-ound spacc chargc clo~ls,” ] . Geo. Res., Vol. 90,

88. R.B. Rent, H.L. Collin, W.C.A. Ilutchinsun, and J.A. C h ; h ~ c v s , “Spacc

chargcs produced by point dischnrgc fnnn trces during a t l , ~ , , , d ~ ~ ~ t ” ~ , ” , ” ] , Atm. Ter. Pl?ys., Vul. 27, p p 67-72, 196.7.

X9, J.A. Clislmers, “Point-dischargc curiiints rhrougli n living trec during a

thondcrstorio,”]. Atm. Ter. Wys., Vol. 24, pp. 1059- 1063, 1962. 90. D. Mulle~-Hillebl-ai~d, “Changc in thc pnth of lightning by ionizing

radiation and spacc chargcs,” Blekirorrische Z. Aas& A, Vol. 83, p p 152-157, 1962.

91. H. Norindcr and I I . Siksna, ‘‘Ionic density of the attnorpheric air inear the ground during tliondcr-stow conditions,” Ark. Genfys., Vol. I, p p 453-472, Dec. 6 1950.

92. B.A. J ~ C O ~ S O I ? and F,.P. Kvidcr, “Elecrroirstic field changes produced by Ihr idn lightning,”J. Atm. Sci.. Vol. . i i , pp. 103-117,lan. 1976.

93. F,.P. Kridev and J.A. Musser, “Manwcll cwicii ts ~ i i idcr thundel-stulma,” I . Geu. Res., Vol. X7, pp. I117J~l1176, Dec. 20 1982.

94. R.B. Stnndler and W.1’. Winn, “Effccts of C W O ~ ~ C on electric ficlda henearh tlilllldCI.stOiills,” Quart. .I. Iloy. Met. Sac., Vol. 105, pp. 285-302, 1979.

95. T. Tali;ihahi arid Y. Sugiooma, “Effect of gmond stiiictiirc t i n iiiipiilsc disch;irgc,” i n Xth lirternational Symposiuiiz on HiRh Vidtagc Eiiginccring Proceedings, Vol. 3 (Yokohams, Japan), pp. 40 1-404, ILL of Jspan,Aug. 23 1993.

96. D.W. Zipsc, “Lightning prorectiun systcms: Advantages and disadvantages," iii I’roceediir~s of the1 Petroleum and C h c m i d Irrdustry TcchnicalCu,*f~.rence(St. I.ouis,MOIISA),p. 343,IEEE,Scpt. 13 1993.

97 .0 . AlconchrlsndB. Thirion, “Study o fa type olcarlystrrsincr ciiiissioii lightning condiictor,” i n WoAshup un Physics of IightninR ((:hamunix, I’l-ance), Fch. 2 1993.

98. A.M. Cassie, ”The efirct of a iadioactivc sowcc 011 the path uf a lightning strdcc,” tcch. ~ p . , Elcctrical Ilescarch Association, Surrey, UK, 1969.

99. J.I;,. Ilobel-ts, “Ionizing radiation and lightning hazards,” Nut., Vol. 21 0, pp. 5’14-515, AI^. 30 1966.

100. j.5. Cheng, “RC circuit tmodel 01 liglmiing b e a m on a very tall structuve,” in Hbctiornqnctic C:oirrpz,nlibilily 1984 Scuexth

3671.3676, jUiy 20 1976.

pp. 6059-6070, J ~ ~ c ~ ~ 30 198s.

JanuaryiFebruaiy 2000 -Vol. 16, No, 1 19

lntematiunal \Vawnrsaw Symposium (Warsaw, I’oland), pp. 183-193, Warsaw Technical University (Warsaw, Poland), June 18 1984.

101, T. Shindu and Y. Aihara, “Shielding theory for upward lightning,”

102, G. Bcno, “USC of ionization in the air for light~~ingp~otection,”Isotiip. Rad. Tech., Vol. 8, pp. 178-1x0, Winter 1970-71.

103. J.R. Gumley, C.G. Invernizzi, M. Khaled, and C.W. WalIhauscn, “Coaxial lightning protectiun for land and marine telecommrmi- cations,” Com. I,,/., Vol. 4, pp. 52-55 , Sept. 1977.

104. J.R. Gomley, C.G. Invernirzi, M Khalctl, and C.W. Wallhausen, “Nuclear lightning protectioii and the new coaxial lightning protection system,” tech. rcp., Nuclear Regulatory Commissiun-in rcpt. “Radioactivity in Cunsumcr Products,” 1978.

105. L. Lefort, P. Bailloz, M. Lefort, and B. Lamhin, “Method and apparatus for protection against lightning,” June 21 1988.

106. L.M. Ball, “The laser lightning rod system: Thunderstorm domesti- cation,”Appl. Opt., Vol. ‘13, pp. 2292-2296, Oct. 1974.

107. AA. Barnes and R.O. Ilerthel, “Survey of laser lightning rod techiii- qucs,” in Inteniatiowal Conference ott Lightning and Static Electricity (Cocoa Rcach, FL), Aug. 21 1991.

108. T. Shinda and S. Sasaki, “Devehpmcnt of laser techndogy for lightning induction,” ,l. Inst. Elect. Eng. Japan, Vol. 1 11, pp. 739-746,

109. S. Uchida, Y. Shimada, C. Yarnnnakn, E. Fujiwara, Y. Izawa, T. Yamanaka, S. Nakai, D. Wang, Z.1. Kawasaki, K. Marsuura, T. Nagai,Y. Sonoi, and N. Shirnokura, ‘‘Laser triggered lightning cxpcriments iii t h e laboratory,” i n 8th Iirtcmational Symposium on lfigh Voltage lhginecring l’xmrcccdings, Vol. 3 (Yokohama, Japan), pp, 313-316, IEE of Japan, Aug. 23 1993.

110. T.P. Chew, “Some aspects of the lightning protcctioii mechanism and radioactive lighming rods,”]. Inst. Hng., illalaysia, Vol. 31, pp. 58-67, June/l)ec. 19x2.

I l l . T.P. Chew, ”The mechanism of lightaingprotection,”Master’s thesis, University of Mnlaya, Faculty of Engineering, Sept. 1979.

112. A.A. Moghissi, P. Paras, M.W. Carter, and 1I.F. Ilarker, “Radioactivity in C O I ~ S U I ~ I products,” tech. rep., Nncleat Rcgiilatory Coinmissinn, 1978.

113. K.P. Iieary,A.Z. Chaherski,S. Gumley, J.K. Gumley, F. Richens,aiid J.H. Mornn, “An cxperimental study of air terininal pcrformance as a function of geomerrical shapcs,” in French-Japanese Workshop (Dcauvillc, France), p p 1-7,July 15 1989.

114. K.P. I-leary,A.Z. Chaberski,S. Gum1ey.J.R. Gumley, F. Richens,aiid J.H. Moran, “An experimental study of ioniring air terminal

Power Engineering Society (Porrland, Oregon), J d y 24 1988.

115. K.P. Hcary,A.Z. Chabcrski,S. Gumley,J.R. Gumlcy, F. Richcns,and J.H. Mol-sn, “An exprrimcnral study of ionizing air rcrmiiiill performance,” IEEE Trans. Power Delivery, Vol. 4, pp. 1173-11 84, Apr. 1989.

116. K.P. Hcary, AZ. Chaherski, F. Richcns, and J.H. Moran, ”An cxpcrimcotal study of corona-ion current uf air terminals,” in 2nd Inrernational Cmrfermce ox Applied Electrostatics (Bcijing, China), pp. 349-354, Nov. 4 1993.

117. K.1’. Heary, A.Z. Chabcrslii, F. Richcns, and J.H. Moran, “Early streamer emissim eiihanccil air tcrininnl QCrfOrCnmCe and zunc of protection,” in I’raceedings ofthc I S Industrial and Commercial Power Systems Anrrual Technical m e (Sr. I’etcrsburg, PL), pp.

118. H. Baatz, “Radioactive isotopes do not improve liglitiiingprotection,”

Trans. Power Delivery, Vol. 8 , pp. 318-324, Jan. 1993.

Sept. 1991.

Elektrotechnide Z. A, Vol. 93, pp, 101-104, Fcb. 1972.

119. R.U. Bent, “Lighmingprotectioii for huildings, rowen and personnel,” Annual 7extile Iirdustry Conference 1986 (Charlotte, NC),

IEEE (avail. from IEEE Service Cent. Cat. Nunher 86CH2320-0. Piscatnway, NJ), May 7 1986.

120, C. Bouquegnean, “The value of radioactive lightning conductors,” Fire Pev., Vol. 172, pp. 28-30, 1984.

121, B.J.C. Burrows, “Rcvicw of alternative systems,” i n Lightning I ’ m tection Seminar Proceedings (Leatherhead, UK), pp. 3.311-7, ERA

122, B.J.C. Burrows, “Thc Franklin rod-An update,” in IEE Conference Publication no. 2.16, pp. 41-46, Itistitutioii of Elcctrical Engineers, UI<, 1984.

123. A. Corvino, “Lightning protection for building structiires,” Elettrifi- caziotie, pp. 493.501, NOV. 1982.

124. R.H. Golde, “Protection of structures against lightning,” Proc. Inst. Elect. Eng.-Cont. Sci.,Vol. 115,pp. 1523-1529, Oct. 1968.

125, P. Leoni, ”Radioactive isotopes do not improve lightiiing caiidiictor performance,” Elcttrificazionc, pp. 227-233, May 1973.

126. W. Rison, “A study of lightning strikes in the vicinity of a radioacrivc preventor,” tech rep., Ncw Mcxico Tech, 1.angmuir Laboratory, Socorro, New Mexico 87801, 1991.

127. W. Rison, “A cumpariron of rhc corona current from a radioactive and noo-radioacrive prevrntur,” tcch rcp., Ncu, Mcxico Tech, Langmuir Laboratory, Socurro, Ncw Mcxico 87801, 1991.

128. H.I.. Suibelaon, “Do lightning arresrers with B radioactive tcmsinittel’ provide a wider field of prutccriun rhan Franklin lightning arresters?,”

129, M.D. Vuela, ”The radioactive lightning cuoducror,” Rev. Electrotec., Vol. 75, pp. 115-120, MaylJunc 1989.

130. M. Belli, M. Cremuncsc, and S. Grcco, “Hcalth implications of the risks connected with die use of americium 24 I for IightiiiiigpTOtCCti”,,,” tech. rep., Istituto Sopcriore di Siinita, Rome, Dec. 10 1975.

13 I, M. Belli, P. Salvadori, E. Sgrilli, and A. Susanna, “Public hcaltb aspects in the use of vadium-226 and americiom-241 i n liglitning rods,” tech rcp., Nuclear Regulatory Commission-in rept. “Radioactivity in

132. G. Bcsscghini and F. Zampioi, “Radioprurection i n the iiistallntioii of lightning-rods cquippcd with radioactive SOLIKCS,’’ C. /:is. Sunitar. Prota. Contro Rad., Vol. 18, pp. 80-87, Jan.-June 1974.

133. PJ. Gillespie, ”Ionizing radiation: A putential lightning hnzard?,” Nat., Vol. 208, pp. 577-578, Nov. 6 1965.

134. P.D. Marliovic and I). Ristic, “Rarlionctivc lightning rad with gamma soiirccs and radiation problem of individuals a i d population,” Tchnika, ~ 0 1 . 3 6 , NO. 9 , p p . 1331-1334, 1981.

D.R. Nikezic and P.D. Markovic, “A Monte Carlo calculation of the eapusure dose due to thc radionctivc lightniiig rod,” Acta Phys. l-len.,

136. G.N. Alcksandrov, G. Bergcr, niid C. Gary, “New investigations in thc

lightning protection of substations,” in CIGRE, pp. 23/13-14, 1994. 137. G. Bcrgcr, “Detcrmination of the inception electric ficld of tlic

lightning upward leader,” in 8th Intentarionnl Symposiim on HiRh Voltage E,igineering Proceediiigs, Vol. 3 (Yokohama, J”pU1). pp. 225-228, IEE of Japan, Aug. 23 1993.

138. G. Bcrger, “Enrlystresmer emission lightning rodconductor,”in 1992 Internotiom1 Aerospace and Cround Co,rferertcc oil Lightning and Static Electricity (Atlanric City, NJ), pp. 38:1-9, Oct. 6 1992.

139. G. Berger, “Formation of the positive leader of long air sparks for various types of rad conductor,” in 22nd lntemational Conference on Ligi,tningP~,,tectiotr (Budapest, Hungary), pp. 43-46, Scpr. 1 1994.

140, G. Bergcr, A. Goldman, B. Senouci, and M. Goldman, ‘%me basic idees to iinpruvc lightning interception.” undated preprint.

T ~ ~ I ~ ~ ~ ~ I o ~ ~ , D K 1 1987.

R ~ V . Electrotec., VOI. 64, pp. ~ ~ ~ - ~ ~ s , J L I I ~ ~ A L I ~ . 1978.

c ~ ~ ~ s ~ I ~ I c ~ Prociucts,” 1978.

vol. 59, N ~ . 1, 91-93,198s.


141. S. Bursteiii and E. Mnriana, “Active and passive lightning protcctian,” Rev. Elcctuotcc., Vol. 76, pp. 143-152, Sept. /Oct. 1990.

142.Y. Aihara,T. Shindo, M. Miki, and?’. Sumki, “Laser-goidcd discharge dimncteristics of lung air p p s and ohservatioil of thc discliargc pvoccsscs,”hlec. hg . ]apan , Vol. 113, p p 66-77, June 15 1993.

143. A.S. Gayvaronsky, A.G. Ovsyannikuv, and J.M. Razhansky, ”A preven- tion o f the long spnrk Icadcr chnnncl turning hy laser rsdiatiun,” in 8th Intematiunal Symposium on High Voltage fingineering I’roce.editq+, Vol. 3 (Yokoliama, Japan), pp. 317-320, 1F.E ofJapan, Aug. 23 1993.

144.J.K. <+reig,l<.F, ~crnslcr,D.P.Murpl~y,R.E.Pechacek,J. I’enn,andM. Raleigh, “l.aser-guided rlectric discharges i n the ntnmosphrre,” i i i

Seventh International Coirfercnce on Gar Discharges and their Applications (Lundun, UK), p p 464-467, Aug. 3 I 1982.

145.J.R. Grcig, D.P. Murphy, 1I.E. I’echacek,M. Ralcigh, E. Lnikio, andS. Hiuvcr , “Pulsed power considerations for lascr guidcd discharges and thcir applications,” in 5th I I’ulsed Power Conference (Arlington, VA), pp, 586.589, June 10,1985.

146. J.R. Cveig, R. Pechscek, M. Raleigh, I.M. Vitkovitsky, J. Hal le, and R. Feruslev, “Interaction uf laser-induced imii7ation with electric fields,” in AlAA Fluid & I’lasina Uynaiizics Coiifcrence (Snowtnaes, CO) , pp. Pap. 80-13R0, July 14 198U.

147. M. Mirohaahi, H. Miyata, C. Sakae, 1’. Sakei, and T. Uchiyama, “Triggering lightning by lascr produced plasma,” in Conference on Lascrsn,,dElcctro-Optics (Ualtimore,MD),pp.310-313,Apl..24 1989.

148. C. Honda, M. Naliarawa, T. Takuma, K. Uchino, K. IMuraoka, P. Kinosliita, 0. Karahira, and M. Akazaki, “A study on laser-induced discharge in atmospheric air (U),” in 8th lntenintionul Symposium on High Voltage Engineeriiig I’roceedings, Vol. 3 (Yokohama, Japan), pp.

149. G. Iinada,’I. Korodii, T. Ohmumo, K. Saiki, K. Masugnta, K.Yatsui,T. Yasuota, S. ‘I‘amagawa, S. Satoh, and T. Cotoh, “l’olsc-power technology and its applications to artificial coiitrol of lightning,” in 8th hrternational Symposium an High Voltage Engineering Proceedings, Vol. 3 (Yokohama, Japan), pp. 309-3 12, IEE of Japan, Aug. 23 1993.

150.M. Ivis,T. Shinkai, K.Araki,andS.YosI,ik;,w~,,“TemporeIdrcayuf 811

ntmosphcnc plasma leader strole for lightning mitiation,” in Procccdirtgs of the Tenth Ixternatiuiral Coiiference on Gas Discharges and their Applications, Vol. 2 (Swansea, UK), pp. 576-579, Sept. 13 1192.

151. R. Itntani, M. Kuho, M. Jinw, G. Nagana, Y. Son& and 7. Nagai, “Fundamental irsrsrcli UII lascr triggcrcd lightning using a new rcchniqoc,” in 8th lnlernational Symposium mi High Voltiige hg imer ixg Proccediqs, Val. 3 (Yokohama, Japan), pp. 301-304, IRE of Japan, Aug. 23 1993.

152. M. Miki, Y. Ailiara, mid T. Shindo, “Development of long gap clischavges giiided by a pulsed CO1 laser,”]. I’hys. 11: Appl . Phys., Vol. 26, pp. 1244-1252, Aug. 14 1993.

1.53. K. Nakamura, T. Suzuki, C. Ymiabe, ;and K. I-loni, “Fundnmcntnl rrsea~cli for lightning trigger experiment hy wing UV lascrs,” Trans.

1.54. 1’. Sbindo, Y. Aihsrs, M. Miki, and T. Suzul<i, “Model experiments uf laser-triggeved lightning,” IEEE Tram. Power Ileliuery, Vol. 8, pp. 311-317,Jan. 199.3.

155. T. Shindu, M. Miki, Y. Aihara, and A. Wada, “A study of laser-trig- gcred lighming-calculafion of plasma generation along a l a m beam,” in 8th Intematiowd Symposium mi High Voltage Engineering Proceedings, Vol. 3 (Yokohama, Japan), pp. 30.5-308, IEE of Japan, Aug. 23 19Y.3.

156. H. Takeno, M. Cho, K. Ohm, S. Nakamoto, and K. Araki, “Basic experiment on r ~ t i i i ~ t i ~ n uf lightning by laser indiiced discliargc i i i

cloud,” iii Xth Intcmatio,ial Symposium on High Voltage Engiirecring

289.212, IEE of japan, A U ~ . 23 1993.

Inst. E I ~ C E % ~ . jnpafr, B, vol. 1 1 3 , pp 126.5-1273, N””. 1993.

Proceedings, Vol . 3 (Yokohama,Japan), pp. 2Y3-296, IW ofJapan, Aug. 23 1993.

157. C. Yiirnahe, K. Nakamova, and K. Honi, “Tcchnology for triggered lightniiigbylascrheam,”OyoButruri,Vol. 61,pp. 718-721, July 1992.

158. I(. Yatsui, I(. Masugata, T. Kuroda,Y. Ohmomo, G. lmada, S. Saroh, T. Goto, K. Ynsuoka, and 1. ‘Tamngawa, “Ureakdown characteristics in atmosphere by T E A ~ C O 2 laser,” i n Proceedings of the Sl’rE-’Ihe lntemational Society fw Optical B:nginceri,ig, Vol. 1810 (Crctc, Grrccc), pp. 671-674, Sept. 21 1992.

159. M. Laan and 1’. Peeter, “The multi-avalanche ~lature of streamer

formation in inharnogeneous ficlds,”]. Phys. D: Appl. I’hys., Vol. 27,

160. I.W. McAllisrer, C.G. Crichton, and E. Bregosbo, “Experimental study 011 the mset of positive curuiia i n atmuspheric ni,” J. Appl. Phys., Vol. 50, pp. 6797-6805, Nov. 1979.

16 I. I.C. Somerville, 0. Farish, and D.J. Tcdford, “Thc influence of atmospheric negative ions on rlie statistical time lag to spark hrcakdown,” in Gaseous Dielectrics Iv -Proceedings of the Fourth lntematianal Symposium on Gaseous Dielectrics (L.G. Chriatophorou and M.O. I’ace, eds.) (Knoxville, TN), pp. 137- 144, I’ergamun Prcss, NewYork, May29 1984.

162. T.E. Allihoiie and D. Dring, “Inflocncc of humidity on thc breakdown o i sphere and rod gaps under impulse voltages of short and long

wavrfrunts,” IEE Proc., Vol. 119, pp. 1417-1422, Sept. 1972. 163. S. Grzybowski atid E.11. Jenkins, ”Estimation of Iightningperfurmnncc

on niodels of 115 kV trniisniissiori liim,” in 8th brternatio?ial Symposium on High Voltage Engineering Pmcmdixgs, Vol. 3 (Yokohama, Japan), pp. 325-328, IEE, of Japan, Aug. 23 1993.

164. W. Kohrmnnn, “Influcnce af wntw vapor on t l ic clectrical hreakdown oiair,”A,rimlen L r Phyrik, Vol. 18, pp. 379-384, 1956.

165. P.N. Mikropoulos and C.A. Stassinopoolos, “Influence of homidity on the breakdown inechaiiism of medium length rod-plane gaps S ~ L C S S C ~ by positive impulse vultngrs,” IBb; Proceedings Sci. Meas. Technol, Vol. 141, pp. 407-417, Sept. 1994.

166. 1I.T. Wntcrs, “Spark breakdown i n non-uniform ficlds,” in Electrical Breakdown of <;uses (J.M. Meek and J.U. Craggs, eds.), pp. 385-532, New York, NY: J ~ h n Wiley & Sons, Ltd., 197X.

167, H. Ryzko, ”The transition from multiple-avalancl,e to single-avalanche mechanism i n the breakdown of air i n n homogcneous ficld,” h k . Fysik, Vol. 25, Nu. 3 5 , pp. 481-507, 1963.

168. H.I. Anis, “Astody of cnrly discharges iii air gaps,” 11 App., Vol. IA-16, pp. 566474, JulylAug. 1980.

169. I.W. McAllirrer and A. Pederscn, “Curma-onset fieldLatrength CBICUI- atims and rlie equivalent radius conccpr,” Arch. Elcktrotechnik, Vol. 64, pp. 43-48. 1981.

170. H. Parekh and K.D. Srivnstaw, ”Effcct of avalnnchc space charge field 011 the calculntion of corona onset voltage,” IEEE Tronr. I?lcctr. Insul., Val. EI-14,No. 4, pp. 181-192, 1979.

171. L.C. Thanh, ”Negative corona in a multiple interacting point-to-plane

pp. 970.978, 1994.

Trans. fnd. ~ p p , % VOI. IA-21, pp. 518-522, Mat.lt\pr. 198.5.

172, M.I. Large and E.T. Piercc, “Thc dcpeiidciicc of poilit-disclinrgc CUII’CII~S on wind as examined by B new cxpcriiiiental approach,” /, Atm. Ter. Phys., Vol. 10, Nu. 516, pp. 251-257, 1957.

173. T.E. Allibonc and D. Dring, ”Influence of radiation on the sparkuvcr of sphcre-piatic gaps stresscd with impulse voltages,” IBH I’rvc., Vol. 121, pp. 759-763, July IY94.

174.Y. Gosha, “~nhancernentofDCposirivestresmercomna inn point-plane gap in air due to addition uf a small ninoiiiir of an clcctl-mcgative pisr”]. l’hys. 1l:Appl. I’hys., Vol. 14, pp. 2035-2046, 1981.


175. J.R. Gumley, ”Lightning inrrrception,” iii Ihctric Eiiergy Coefcre?ice 1989: b:lectrical Sovices for Uuildixgs. Preprints of Papers (Sydney, NSW, Aucmlia), pp. 136-141, Natl. Cmmn. Electr. Energy Coll. Elcctr.

176. J.R. Gumley, “Lightning I n t e r c e p t i m a n d the uplcader,” in L2ird International Conference on Lightaing I’rotectiun (Budapcst, Hungary), Sept. 19 1994.

177. C.R. Moorc, J.A. Mathis, and W. Ilison, “Report on thc C X ~ O S U Y C of eighteen air terminnls along die crest of thc Magdalena inowitailis iii

New Mexico doring the 1994 thondel-storm BCBSOII,’’ tech. rep., New Mexico l e c h , Liungmoir Laboratory, Sucorm, NM 8780 I ,1994.

178. A. Mumknmi, “1.ightning protection zones for rocket launch complexes,” rcch. rep., Air Force, 1993.

179. G. Bergrr, B. Scnuuci, A. Goldman, a n d M. Goldman, “A physical approach for lightning protection,” in I X International Cmifwencs on

G a Uischnrge and their Applicntiorrs (Venc,.ia, Italy), pp. 435~438, Sept. I9 1988.

180. G. Hnrtniann and I. Gallimherti, “lk influence of mcrastnblc molecdes on the strcanics progression,” J. Phys. D: A/$. I’hys., Vol. 8, pp. 670-680, 1975.

181. J.J. Lowke, “Theory of clecrrical breakdown in air-The rule of inictastahle oxygen ~~~olccuIcs,” J. I’hys. D: Appl. Phys., Vol. 25, pp. 202-210, 19Y2.

182. N.L. Allen ancl D. Diiiig, “Variation LII corona forinatim mrlcr repetitive itiipiilse conditims,” ill Fourth lnteniatiiinui Symposium on

183. O.P.A. Domcns, 1’. Gilbert, B. IIutnlcl-, and G. Riquel, “Perfurmancc of a 16.7 1x1 air rod-planc gap under a ncgntivc witching impulse,"/. I’hys. 1): AppI. I’hys., Vol. 27, pp. 2379-2387, 1994.

Geo. Rcr., Val. 76, pp. ,5799-5806, Aug. 20 1971. 185. F,,O. Selim and 1l.T. Waters, “Space charge modclling in impulse

c ~ m m , ’ ’ in Fourth Interitational .Symposium on High Voltage

I 86, L. ‘I’hiunc, “Tlic dielectric strength uf lnigc air insulation,” in Su~ges in High- Vo!tqe Netwarks (K. Ragaller, d), pp, 165~205, Plcnuin Press, NewYork, 1980.

187. W. lhrkowski, “Study of the long sliding spark,” Archiwuni Elektrw technitii, Vol. 3 0 , No. 1, pp. 279-295, 1981.

I n n . Y. Watanahe, “Switching surge flashover cliaractcristics of extremely long air gaps,” IEEE Trans. I’owerApp. Sys., Vol. PAS-86, pp. 933-936, Aug. 1967.

1 89. C. Gary, “l’aramercrs for H.V. tcstingofverioos form of air teriniiiatimis.”

190, 1I.H. Gulde, “Thc attriictive effect of a lightning conductor,” J. Inst.

191. T. Suzuki, K. Miyalie, atid I. Kishizimn, ”Study on expcrimcntal .am. PowerApp. Syr., Vol. 100,

pp. 1703.1711, ~ p r . 1981. 192. les Reiiardieres Group, ”Part 4. Effects of pre-existing space charge on

positive discharge development,” l i C Proc. A, Vol. 133, pp. 469-483, Oct. 1986.

193. L. Dcllrra and E. Garhgnati, “Shielding failure Cvahtioll: Application of rlie leader progression n i ~ d ~ l , ” in lnte,rrutionai Conference oil 1.iEharig mrd I’ower Systems (I.ondon, England), pp. 31-36, IFE of

London, Eiiglaird and New York, NY, Junc 51984. 194. J. Dupuy, “Negarivc diachargc in air,” i n I’roceedings of the 4th

Japan-Francs Workshop on Lightning (Kaonz;iwa, Japan), pp. 19-24, Nor. I 1 1991.

195. C. Cnry atid U. Ilutzlcr, “Simulntim a i lightning artachmcnr to

cartlicd stn&rcs,” I<m. Gen. Electr., pp. 18-24, Mar. 1989.

E, ,~. , 0~1.24 1989.

~ i ~ / > voltage ~,tgixccri,tg (~t i le t l s , (;recce), p. 41.01, Sept. F 1983.

184. C.T. PhelpS, “l’icld-cnhnnced p~Op>~g~tiOtI U f COrOlla StK:llllelS,” 1.

~ ~ ~ i ~ ~ ~ ~ i ~ ~ (~ti,~,,~, G ~ ~ ~ ~ ~ ) , Sept. s 1983.

LIlld:& pl’epl‘i~lt.

Elecl.Bvg.,V~I. Y,pp. 212-213, May IY63.

simulation of lightning strokes,” I

196. C. Gary, 1%. Hutzler, D. Crietisc, (;. Dragan, R. Enache, and B. P o p , “1,nharntory aspects regarding the iipward positive disclisrge due to negative lightning,” Rev. Ruumaine Sci. Tech., Eldctrotechniqae et Bnergetiquc, Vol. 34, pp. 363-377, July-Sept. 198Y.

197. E. ~;ackenI~acl~,~“Lightning discharge siiiiulation in thc Inburarory,” Bull. Schweizeyischen Clektrotechnischeir Vcrcins ~3 des Verbandas , ~ c b w ~ h e r i s c h c r E i e k t r i z i t a ~ t ~ w ~ ~ k e , Vol. 80, pp. 239-247, Mar. 4 1989.

198. B. H u w h , “Comp on uf lightning and loiig spark,” Rev. Gen. Electr., pp. 12-17> Mar. 19x9.

199, B. Hutrlcr, “Lightningsimulariun,”BtrlI. Dir. Etud. Rech. U, pp. 3 1 ~ 4 0 ,

200. B. Jones, “Switching surges and ail- insulation,” Phil. Treans. Roy. Soc. Lond, A, Vol. 275, pp. 165-1x0, 1973.

201. S. Larigaldie, C. I.ahaunr, and J,P, Morcan, ”Lightning leader laboratory s i d a t i m i hy iiiraiis of rectilincar surface discharges,”]. AlJpl. I’hys., Vol. 52, pp. 71 14-7120, Dec. 1981.

202. E.T. I’iercc, “Natural lightning paramerers and their sinnhtion in 1abol.atory tests,” in 1 Y 75 Confererrce ox Lightnixgand Static Electricily (Abingdon, Uerks. UK), p, 13, Apt. 14 1975.

203.1‘. Suzuki, K. Miyake, and 1. Kishizima, “1,ight~~ingsrrokes to traiisiiiission lines a i d substations. 1. Lahorarory experiments of lightning strokes,” tcch. rcp., Cenrral Research Institution of Electric I’ower Industries, Tokyo,

204. S. Sspur nod W. Thrkowski, “Laburatory corroburariun of the relaxntim rhcory of the lightning s t ~ p p d Icadcr,” Archiwum Elektrotechniki, Vol. 17, No. 2, pp. 405-407, 1968.

205. M.A. Urnan, “A comparison of natural lightning and tlic long laboratory spark with applicatiun to lightning tesring,” tech rep., Wcstinghouse Research Labs, I’ittshurgh, P1\, Aug. 11970.

206. C.F. Wagner and A.R. Hlleman, “Mechanism of brcakduwn of

scpt. 1988.

~ a ~ ~ ~ ~ , uec. I 1978.

iaboratary g a l , s , x ’ ~ r r a u n s . AWL I , ,SI . EIC. ~ n g . , 111; VOI. no, pp, 604-622, Oct. 1961.

207. G. Carmra and L. Thione, “Switching surge strcngth of large air gsps: A physical approach,” IEBE Traris. Power App. Sys., Vol. PAS-95, pp. 512-524, Mar. /Apr. 1976.

208. A. Carrus, “An inducrancc u n the Marx generator fail branch-Ncw technique for high efficiency lahoratory reproduction of sliort time to half valuc lightning impnlses,” I E E E Trans. Power Delivery, Vol. 4, pp.

209. U.W. Clifford, K.E. Crouch, and E.H. Schultc, “Lightning sinidation and testing,” IEEE Trans. Blectromag, Comp., Vol. 24, pp. 209-224, May 1982.

210, 0. lhrish, “Acting OII impulse: Applicaiions of IIV pulsc tcchnolagy,” Efts. Sci. Ed.]., pp. 277-286, Dcc. 1994.

21 1. ILJ. Fisher and M.A. U m w , “Simulation fidelity in lightning penetration studies,” tech. rep., US Dcpartiiiciit of Encrgy, Fch. 1 1990.

212. A.H. Pautan, L. Uitker, and R.L. Gardncr, “Natural lightning study 198.5,”rech. rcp.,AirForccWeapoiisLab., KirtlandAl:B,Dcc. 11986.

2 13. W. Schufft and W. Schvader, “New Marx gemrator for the simulatim of lightning irnpulsc voltages and c~~rrc i i t s ,” in 8th International Syniporinm on lligh Voltage Engitieering Proceedings, Vol. 3 (Yokohama, Japan), pp, 453.456, IEE of Japan, Aug. 23 1993.

2 14. ESI . Schulte, “Updating thc McAir Lig1,tningSimulatioll Laboratory,” rcch. ~ p . , McDonriell Aircraft Cumpnny, St. Louis, MO, 1983.

21.5. E.H. Schulti., ‘“lhc Mcnonnel Aircrait Coinpany Lightning Simolation I.abaratary,”]. Envimn. Sci., Vol. 22, pp. 22-27, May 1979.

216, L.C. Walk” ancl J.L. Hcbcrt, “Lightning simulation facilitics in the Unired States and Europe,” in 23rdAeror/iace Sciences Meeting (?,,a, NV),AIAA,Jan. 14 198.5.

217. W. Zischank, “Simulation of lightning discharges by direct strikes,” Elektrotechnischc Z., Vol. 105, pp. 12-17, Jan. 1984.

Y O - Y ~ , J ~ ~ ~ ~ . 1989.


2 18. D. Graf, "Calculation of curuiia disclrnrges iii point tu plnnc gaps," in Third Intcnratiuna/ Symposium on High Voltage Eirgisccring (Milan, Italy), pS3.04, Aug. 23 1979.

21 9. L.E. Klinc, "Cornna cloud mode l predictions of switching siirge flnshovcr vultngcs YS. electrode geometry," in I I'ower Engi,mritig Sociely SuinmerMectbig 1976 (Portland, Oregon .P76345-9,July 18 1976.

220. L.E. Klinc, "Monte Cal-lo study oi imimtion sone electron kinctics iii negative pin-plane coronas in arinasplwic air,"]. Appl. I'hys., Vol. SX, pp. 3715-3719, Nov. LS 1985.

22 1. M.M.A. Salama, H. I'arekh, nod K.D. Srivastava, "Curomt iiiccptitin ondcr switcliingsurge for puint- to-phe long gaps,"]. Appl. Phyr., Vol. 47, pp. 2915-2917, July 1976.

222. M.C. Wnng and E.E. Konliardr, "Streamer dynamics," Phys. Rev. A, Vol. 42, pp. 2366-2373, Aug. IS 1990.

223. G.A. Ilawson and W.P. Wion, "Amodcl fol.stl-eamrrpl.opagatioii," %. Phys., Vol. 183, p p 159~171, 1965.

224. I). I)omcos, A. Gilxrt, J. D U Q U ~ , and 11. Mutzler, "Propagation of the

positive streamer-leader i n a 16.7 111 rod-plane gap," J. Phys. D: Appl, Phys.,Vol.24,pp. 1748~17.57, 1991.

225. M.1. Dyakonov and V.Y. Kachoruvskii, "Srrcnmcr dischnrgr in a hornogciieous field," Suv. Phys. ]ETl', Vol. 68, pp. 1070- 1074, M a y 1989.

226. 11. IHiitzler and D. Hurzlcr-l1arre, "Leader propng;nion iiindel for pl.cdetcl.nlinati~,,, of switching surgc flashover volragc of largc air g:ips,"

'Iklnlls. I'ower AfJ/J. SyS., Vol. PAS-97, pp. 1087-1096, JulylAug. 1978.

227. S. I.arigaldic, A. Roussnud, and U. Jccko, "Mcchaoisms of high-currcnt piilses i n lightning and lung-spark stepped leaders," J. A/>/,/. I'hys., Vol. 72, pp. 1729~1739, Scpt. 1 1992.

228. W.B.I. Mnicr, A. Kndish, C.D. Surhcrlnnd, ;and 1R.T. Rohiscoc, "A distrihutcd paranleter wire nodc cl for traiisicnt clectrical discliargcs," J. Appl. I'hyr., Vd, 67, pp, 7228-7239,June 1.5 1990.

229. E. Maiodc, "Thc mcchanisin ofsparkhieakdown in air at ato~inphcric ~ E S S U I Y between a pwitivc point and plane. 1. Ilxpcrimcnral: Natiire of the srl'emer tsack,"]. Appl. Pbys., Vol. 46, pp. 2005-20 1.5, May 1975.

230. F.. Marodc, '"l'he meclisnism of spark breakdown in air at atmusphcric pressure berwccn a pusitivc point and plane. 11. 'Tlicurctical: Compurer siiiidiitioti o f rlic strcatncr tvaclt," 1. Appl. Phys., Vol. 46, pi'. 2016-2020, May 1971.

23 1. E. Nasscr and M. Heisdel-, "Marhcmntical-pl~y~i~~l nod cl oi the streamer in oumnnifurm ficlds,"]. Appl. Phys., Vol. 45, pp. 1396-3401, Aug. 1994.

232. W. Reinioghaiw, "Calculntiun of stveatiiers i n gascous disclmges,"J.

Phys. D:App/. I'hys.,Vol. 6, pp. 1486-1493, 1973. 233. M.M.A. Salama, H. I'arelih, and K.D. Srivastava, "Model for

switching sorgc breakdown of a poini~tu-planc air gap,'J. Appl. I'hys., Vol. 47, pp. 4426-4429, Oct. 1976.

234. 1.. Dcllarn and E. (;avbagnati, "Lightning stroke siniiilatiun by nicms of rlie leadcl- progression model. Part I: Dcscriptioa of the niodcl and evaluation of cx~usure af free-standing structuxs," I Delivery, Vol. 5 ,No .4 ,pp . 2009-2017, 1990.

2.3s. L. Dcllcra and E. Garhagnati, "Ovcrhcnd power line lightning prorectiun," tech. rep., Ente Naziuilalc pcr I'Energin Elettrica, Milan (Italy), 1990.

236. W. Dcrrz and V. Fritsch, "l)etermioatiun of areas of tmisniissiun r u ~ e r s attracting lightning strokes,'' Elcktrichertuo, Vol. 68, pp. 389~394, June 9 1969.

237. G Dragan and M. Ungurcanu, V i e striking distmcc of ligliining," Re". RoamaiaeSci. Tech., Elcctl-otcchniqne et Cnergetiqoc, Vol. 27, pp. 191-197,Apr.-June 1982.

238. A.J. lirikssun, "Tlic incidencc of liglitning strikes to power lincs,(I IEEE 'l'ra?zs. Power Delivery, Vol. PWKD-2, pp. 659.870, July 1987.

239. I.A.M. Rirk, "Modeling of lighming incident to tall structures, Part 11. Applicatian," in IEEE Wintar I'ower Engineering Society Meeting (Columhus, Ohio), pp. 93 WM 081-0, IEEE,Jan. 31 1993.

240. F.A.M. Rizk, "A model fur switching impulse leader inception and breakdown of long air-gaps,'' IEEE Tram. Powar Delivery, Vol. 4, pp, 596-606, Jan. 1989.

241. F.A.M. Rirk, "Modeling of lightning incidence to ta l l structures, Part Winter Power Engineering Society Meeting

(Columbus, Ohio), pp, 93 WM 082-8, IEEE, Jan. 31 1993. 242. F.A.M. Rizli, "Modeling of tlansmission line cxposure to direcr

Iighmingstri~kcs,"IEEE Trans. Power Delivery, Vol. 5 , pp. 1983-1997, Nov. 1990.

243. F.A.M. Rizk, "Critical switching iinpiilsc breakdown of long bundle- conductor gaps," in I Winter Power Meeting (New York, New

strength of air insulation: Leader inception criterim,"Ib:EE Traits. Power Delivery, Vol. 4, pp. 2187-2195, Oct. 1989.

245.J.ll. Stahmann, "Launch pad lightning prutectiuii effcctiveness," tech.

rimal Aeronautics and Space Arlministration, Aug. I 1991. Iiuin, A.C. Licw, and C.M. Wong, " C ~ ~ i ~ p u t e r siiiiulatim of the tages a i t h e lightning protection mechanism," IEEB Trirns.

Power App. Tys., Vol. PAS-101, pp, 4370-4377, Nov. 1982. 247. A. Pederseo, "Calcolation of spark breakdown or corona starting

vultagcs in nonimiform ficlds," IEEE Trarts. Power App. Sys., Vol.

248. t l . Reather, " ' l l ~ c dcvclopmcnt of an electron avalaiiclie into n spark chanocl,"Z. Phys., Vol. 112, pp. 464-489, 1939.

249. R.E. Orville, "l'horographs of a ~ O E C lightning flash," Sci., Vol. 162, pp. 666-667, Nov. K 1968.

250. M.V. Piepgrass, E.P. Krider, and C.B. Moorc, "Lightning and surface rainfdl during Florida thimilerstorms," 1. Geo. Res., Vol. 87, pp. 11193-Il201, Ucc. 20 1982.

25 I . C.B. Moorc, "Prcliminsry report ULI the 1993 exposurc of commercial ESE air terminals mi Suuth Baldy Pcak i n New Mexico," tech. ~ p . , New Mexico Tech, Lsngmoir Laboratory, Soccora, NM, 1993.

252. I(. Honi and H. Saliuwno, "Observation of final jump of the discliarge in the cxpcrinicnr of arrificially triggered lightning," IEEE Trans. Power App. Sys.,Vul. P A S ~ l 0 4 , No. 10, pp. 2910-2917, 1985.

253. J.R. Iippcrt, "Evalilation uf rocket triggcrcd lightning for research snd developntenr," tech. rep., Air Furcc lnstitutc of Tcch., Wright-Pattcrson AFII, 1981.

254. K. Nakamura, I(. Honi, and S. Aiba, "Discharge CLITRII~S in t h e euperinicnt of artificially triggered lighming for winter thunderclouds," i n I989 Intcnzntional Symposium on Uectromagnetic Compatibility (Nngoya,Jspan), pp. 664-669, I (cat. 11. 89TH0276-6, IEEEService Center, I'iscnfawsy, NJ), Scpt. 8 1989.

255. M. Ilrook, G. Armstrong, R.P.H. Winder, U. Vonnegut, and C.ll. Moove, "Avrificial initiatim of lightning disclinrges,"J. Geu. Res., Vol. 66, pp. 3967-3969, Nov. 1961.

2.56. A. I~lierl, "Charncteristics of triggered lightning," in ICLP '8.5: 18th International Canfenttce ow Lightning I'rutectiorr (Munich, Germany),

257. (;.I+. Schnerzrr and R.J. Fisher, "Sandia transportable triggered lightning instrunientation facility," tech. rep., National Aeronautics and Space Administtation, Aug. I , 1991.

258. R.1'. Picux, C.H. Gary, 8.1'. Hutnler,A.R. Eybert-Rcrard,l'.L. Hubcrt, A.C. Mccstcrs, P.H. Pcrrood, J.!-I. Ilamelio, and J.M. Person,

IPAS-86, pp. 200.206, P ~ I , . 1967.

pp. 11-18,Sept. 16 198.5.


“Research on artificially rriggered lighrniog i n Prancc,” IEEE Tram. PowerA[,p. Sys., Vol. PAS-97, pp. 725-733, MaylJunc 1978.

259. X. Lin, C. Wang, Y. Zhaug, Q. Xiao, I). Wang, Z. Zhou, and C. Guo, “Enperimenr of artificially triggering lightning in China,” J . Gea. Res., Vol. 99, pp. 10727-1073 I , May20 1994.

260. A. Eybert-Berard, U. Thirion, J.P. Bcrlnnrlis, B. Undor, and C. Gary, “In situ lightning rod tcsts and analysis,” i n Lighttting n d Mou,rtaixr (Chatnunix Nonr-Blanc, France), Socictc dcs Electricians et Elcctraniciers et Club Alpin Francais, June 6 1994.

261. D. Mullcr-Hiliebrand, ‘“l’he protection of hauscs by lightning c u m &mors-An historical review,” J. Frm. Imt. , Vol. 274, pp, 34~54, 1962.

262. G. Bcrgcr, “Applications of t h e electrogeoinctric model to Fraiiklin and ESE lightning rods,” in Lightning Protectioii Workshop (Hobarc Standards, Australia), Nov. 12 1992.

263. J.F. Shiplcy, “Thc protcctian of structures against lightning,” I. I m t . Elec. Erg. I , Vol. YO, pp. 501~523, 1943.

264. P.G. Woalncr, “Calculating lightning protcctian zones,” tech. rep., Tlic MlTllE Corporation, 7525 Colshire Drirc, Mclean, VA, 1990.

265. R. Rrict, “Tlicory of ground-based lighming protcctiun systcms,” i n Euro EM Conference (Bordeaux, b’mncc), J L ~ C I 1994.

266. S. Srpor, “Lightning protection zoncs,)I Arch. Elektrotechniki, Warsaw, I’uland,Vol. 28, No. 3, pp. S61-577, 1979.

267. U.1’. McKae, J.I.. Dirsrtiilarf, and GT. Glasson, “A )new theury of shielding against direct lightning strokes based on recent lightning discharge data,” in Sth International Conference on Gas Discharger

268. V.A. Kurilav, ‘ T h e orientatioii of lightning discharges on earthed objects,” Izv. Vys~h. Uchcbn. Zaucd. I(mrs., pp. 24-30, Nov. 1978.

269. C. Gary, S. Hurduhctio, and G. Uragm, ”llownward negative lightning and high voltage ouerhend liuc surfacc gd ienrs ,” in 21st lnteriiatiunal Cmrferdnce 072 Lightniiig Protection (Berlin, Gelmnny), pp. 49-5.3, Sept. 21 1992.

(Livcrpaol, UK), pp. 108-111, Sept. 11 1978.

270. M.N.S. Immclman, “l’oint-dischargr CLITI’CII~S during thundevsrorms,” Phi. Mag., Vol. 25, pp. 159-163,Jan. 1938.

271. E. Foriics and P. Ortiz, “Kadioactive lightning rode, static cliiniiiators, and otlier radioactive devices,” tech. rcp., Nuclcar Regulatory Commission-in rept. “Radioacrivity in Cunsumer I’roducts,” 1978.

272. J.A. Ihbbing, A.W. IHaason, and P.F. Little, “Simulated lightning attachments to aircraft skins,” i n 5th lnternntitinnl Conference on Gus Discharges (Liverpool, IIK), pp, 289-292, Sept. 11 1978.

273. A. Kern, “Simulation and measurement uf melting effects on metal sheets causcd hg direct lightning strikes,” tech. rep., National Aeronautics and Spacc Adminisrration, I99 1.

274. J. Phillpott and T.E. James, “Simulation of lightning mikes to aircraft,” rech. rep., UKABA, Abingdon, Berks. UK, May I 1971.

275. E.H. Scholtc, “Lightning damage ~ i ~ e c l ~ ~ n i s ~ ~ ~ s and siiiiiilacion techn- iqucs,”]. Bnviro,~. Sci., Val. 23, pp. 13-17, May 19x0.

276. R.C. Twmney, “Effects uf lahoratory simulatcd precipitation iratic electricity atid swept strokc lightning 011 aircraft windshield sobsystems,” tech. rep., Douglas Aircraft Company, Long Beach, CA, July I 1976.

277. D.M. Leitc, “Expcricncc of radioactive lightning aerials in field and in lahorarories,” in Conference l’mcecdings. IC12 ‘85: 18th International Conference on Lightniiig Protcctimt (Munich, Germany), pp. 127-133, VDE-Vcrlag, Bcrlm, West Germany, Sept. 6 1985.

278. A.C. Liew, “Lightning dainagc and problcrns cxpcricnced i n some structorc~l~gstetlIs in Singapore,” i n Lightning Protection 92: Buildings, Stnicturcs and Electronic 1:quipmeiit Conference and Erhibirioti (l.earherBead, UK), pp. 7.211-8, ERA Tcchoolugy, Junc 23 1992.

279. A.M. Mousa and K.D. Srivastava, “Shielding of tall structures against direct lightning strokes," i n Canadian Coitfcrcnce mi Electrical m d ConipLIterEiisinecri~~ (Vaocouvcr, R.C., Canada), pp. 342-352, Nov. 3 1 9 8 ~ .

Correction

The term “LC Shield,” wed i n the NovemberiDecember 19YY feature, “Technical Advances in the Undergroutid Me- dium-Voltagc Cable Specification of the Largest Invcs- tor-Owned Utilities in the US.,” is a registered trademark owned by the Pirelli Cable Co. In the nse of the term in the

section on Metallic Shielding and the section on Protective Jackets, thcre should have been a trademark symbol (0) following the term whenever it occurrcd with a footnote indi- cating that it is a l’irelli trade mark. The editors regret that this omission occurred.


Documents

Early streamer Emission Lightning Protection · PDF fileEarly streamer Emission Lightning Protection Systems: An Overview Key Words: Air terminal, ... Corona discharge-A localized,