Sie sind auf Seite 1von 6

V International Symposium

on Lightning Protection
17'' - 21" May, 1999 Sao Paulo - Brazil

EXPERIENCE IN THE DESIGN OF EXTE~AL PROTECTION SYSTEMS


AGAINST LIGH1.]NING
Ing. Patricia Arnera Ing. Julieta Vemieri Ing. Beatriz Barbieri
IITREE - UNLP IITREE - UNLP IITREE - UNLP

IITREE- UNLP (Universidad Nacional de La Plata) - 4~y 116 (1900) La Plata -ARGENTINA
TE - Fax: +t54 221 4836640 / 4837017 I 4250804
E-mail: iitree@volta.ing.unlp.edu.ar- http:\\~.iitree.ing.unlp.edu.ar

Abstract - This paper describes the experience acquired local regulations that oblige the installation of LPS. Only in
during the revision and determination of external Buenos f.ires
City there is an obligatory regulation
lightning protection systems in some refinery plants, in (N" 1411-PGFOC-98) that demands protection by installing
Argentina. With this purpose a program, developed by LPS in accordance with !RAM 2184-1 and 2184-1-1
IITREE, based on the Monte Carlo statistical technique, standards! . .
was used, among other things, to determine in a period of i
time the frequency of lightning flashes to structures. Cousiderihg the fact that !RAM standards are based on IEC
International and American standards were used as standardsJ IEC 1024-1 and its sections (IEC 1024-1-1, IEC
reference. 61024-1-t [5)), as fur as the American NFPA-780 [6], were
used as cip:rent standards of reference.
1 INTRODUCTION i
2.2 Z<1nes of protection
In Argentina, nowadays and during the last few years lEC-1024}-l and NFPA-780 standards alter from protective
people, in a less scale politic leaders and important angle, and rolling sphere criteria when assessing proper
industries, concern about damage that many locations for air terminals.
technologies and current way of life are causing to the
environment Among industries some refineries are not IEC defuies values for protective angle, and rolling sphere
only concerned about real risk, but loss of social for each \ corresponding protection level and regarding
prestige. One agent that could cause high damage structure peight, as it can be seen in Table l, where h is
(material and life losses) is the lightning phenomena, height of the structure to be protected in metres, a protective
especially in such plants where flammable products are angle and 1R rolling sphere radius in metres.
being manipulated and processed.
As you can
see in this table, for structures higher than 20
metres, depending on the desired level of protection,
2 REVISION OF CURRENT STANDARDS protective angle method is not applicable. On the other hand,
rolling sphere method is always · applicable, no matter
2.1 Current standards as reference structure size, height, shape, etc.

Standards usually provide information concerning Table I: Positioning of air-termination according to the protection levels
(!EC 1024-1).
design, construction and materials of Lightning
Protection Systems (LPS), however they say nothing
h(m) 20 30 45 60
about necessity of installation of an LPS. Standards of PROTECTION
application in Argentina are: !RAM 2184-1 (1996) [l) LEVEL
R(m) a,(O) a,(O) cx10) a,(O)
and !RAM 2184-1-1 (1997) [2), with the respective
modifications and all complementary !RAM standards. I 20 25 * * *
Mentioned standards are based on, and equal in
contents, IEC 1024-1 (1990) [3) and IEC 1024-1-1 II 30 35 25 * *
(1993) [4) international standards. III 45 45 35 25 *
Generally in Argentina and especially in territories IV 60 55 45 35 25
where the refineries in study are located, there are no * In these cases only apply rolling sphere.
285
As the protective angle method establishes, the zone of 3.2 Assessment of required efficiency for LPS designs
protection forms a cone having an apex at the highest
point of the air terminal, with walls forming an angle Once structure classification has been adopted, section 3.1,
from the vertical. we assigned a recommended protection level for each type
of structure. The pwpose of selecting a protection level is to
As the rolling sphere method determines, the zone of
reduce, below the maximum tolerable level, the risk of
protection includes the space not intruded by a rolling
damage by direct lightning flash to a structure, or to a
sphere when it lays tangent to earth and rests against a
volume to be protected.
lightning protection terminal.
Applying IRAM 2841-1-1 standard (based on IEC 1024-1-
1), we classified refineries as structures dangerous to their
3 ADOPTED CRITERIA FOR PROTECTION surroundings where the effects of lightning could be fire and
explosion in the plant and its surroundings.
3.1 Classification of typical installations Applying NFP A 780 standard we can classify some of the
typical refinery installations as structures containing
In order to determine external protection sy!\(ems flammable vapours, flammable gases, or liquids that can
against lightning, it was necessary to classify different give offflammable vapours.
existing structures into representative types, and adopt
some criteria applicable to protection of each one. In order to determine the proper protection level, we
calculate the required efficiency Ee of the LPS, with the
I
The following structure classification was done: following equation:
TYPE 1) buildings in administrative areas, 'friE 2)
buildings immerse in process plants, TYPE 3) &=l- Ne (1)
recipients containing flanunable vapours, flanujmble Nd
gases, or liquids that can give off flammable vapours
Where Nd is the average annual frequency of lightning
(tanks and pools).
flashes to the structures and Ne is the maximum accepted
annual frequency of lightning flashes which can cause
TYPE 1) Buildings located in non-dangerous lireas, damage to each type of structures, estimated in accordance
usually are made of concrete, or in some case$I are with IRAM. 2841-1-1 and ENV 61024-1 (European
made of metal sheets. Almost all of them cait be standard) as follows:
classified as common structures. i
5.5·10- 3
TYPE 2) In process plants yon may accept the pre5ence Ne= C [flashes/year] (2)
of explosive atmosphere, because of the fl~able
substances being manipulated. Strippers, rebo/lers, Where C was calculate with the following equation:
columns, reactors, compressors, charge heaters,
coolers, condensers, refrigeration towers, motors) etc. (3)
are typical equipment in these areas. Also electric
substations and other buildings can be found iin a C, : coefficient that evaluates type of construction of the
process area, usually, not containing flamriiable structure.
substances. C3 : coefficient that evaluates structure contents.
TYPE 3) Storage recipients, containing petroleunt and C4 : coefficient that evaluates structure occupancy.
petroleum products are made of metal and the great
Cs : coefficient that evaluates consequences of a direct
majority are thick enough not to be punctured jby a stroke to the structure in the surroundings.
direct strike and are normally well grounded sq that
they do not require lightning protection. However, in Nd is calculated as a product of the local ground stroke
some of the refineries being analysed metallic tlmks, density Ng and the equivalent collection area Ae of the
•· I
although having enough metal thickness no tp be structure:
punctured, suffer from lack of maintenance al)d in
some cases show holes where flanunable vapoutj; can -6
Nd=C ·Ng·Ae·IO (4)
give off. Consequently, lightning protection will be
1
required in such tanks. In addition, usual open-air Pools Where c, is an environmental coefficient taking into.
~=::::!, flanunable vapour will require lighf"ing account relative location of the structure.

286
Table 2: Parameter values used to calculate Efficiency Ee, for. each 3.3 Adopted protection method
type of structure.

Common Buildings Tanks Pools Considering , clients preoccupation in relation with


buildings in process enviromuental impact and social consequences, protection
plants levels selected were more severe, when possible, than the
ITYPE l' ITYPE2) (TYPE3 1Yl'E3) smaller ones 'coming from the following equation:
Ae lm"l 2860 2860 10936 2534
• Ne
CI I 0.25 I 0.25 E;;,:&=1-- (7)
Nd [fish 0.0088 0.0022 0.0385 0.0022 i Nd
I
I yearl . Rolling sphere method was employed to design alternative
C2 I 1 0.5 3 LPSs, and the following protection levels were proposed for
C3 0.5 1 3 3 the previous I'structure classification:
C4 3 1 0.5 0.5
cs 1 10 10 10 Protection level I, and additional protection measures
(R = 20 lm), were applied to Type 3 structures: tanks and
c 1.5 10 7.5 45
pools contaiuingflammable vapours, flammable gases,
Nc[flsh 0.0037 0.0006 0.0007 0.0001 I
or liqui(ls that con give offflammable vapours.
I vear]
Protecdon level II, R = 30 m, was applied to Type 2
structurl:s: buildings located in process plants, such as
Table 3: Calculated efficiency Ee, and efficiency E corresponding electric I substations, control rooms, dressing rooms,
with protection levels. refrigeration towers, etc.
I
Common Buildings Tanks Pools Protecfion level IIL R = 45 m: was applied to Type 1
buildings struc~s: common buildings located in administrative
areas ,and other non-dangerous areas, such as
managtjment, laboratories, medicines, etc.
Ee
E<!Ec 3.4 An~ysis of the selected protection level
Level I*
Comparing iboth standards, NFPA 780 defines the zone of
protection for common stnictures with a rolling sphere
Expressions applied to obtain Ae, given in the having a rlidius of 46 metres (150 ft) in accordance with
standard, are: protection level III as lEC 1024-1-1 defines. This is a good
Rectangnlar area Ae = ab+ 6h (a+b) + 9 11 h2 (5) reason for c~oosing level III for Type 3 structures.
I

Roundarea Ae = 11 (¢12 + 3h) 2


(6) In the samd way, NFPA 780 defines the zone of protection
for structu1-es contaiuing flammable vapours, flammable
Where a and b are the object length and width gases, or liquids that can give offflammable vapours with a
respectively, h is the object height, and ip is the circle rolling sph~re having a radius of 30 metres (100 ft), in
radius. accordance:with protection level II as lEC 1024-1-1 defines.
Trying to satisfy client concern, about social impact, level I
Typical structure dimensions assumed are as it follows:
was recommended for such structures (Type 3).
TYPE 1) common building: a = 20 m, b = 20 m
andh=6m.
4 APPLICATION OF THE ANALYSIS TOOL
TYPE 2) building in process plants: a = 20 m,
b = 20 mandh = 6 m.
4.1 The computer program
TYPE 3) tanks: ¢= 23 m and h = 12 m.
A computer program developed by IITREE called
TYPE 3) open-air pools: a = 56 m, b = 35 m and
BLINSUB, assists in determining the objects beiug struck
h = 1 m for railing height rounding the pool.
when a number of lightning flashes moving downward a
Quantities indicated in Tables 3 and 4, were used to region are simulated. It employs the Monte Carlo statistical
calculate the required efficiency for each type of technique to select lightning by means of an external file
structure classified as we proposed in section 3.1. with an empirical distribution for current amplitude, and
chooses flash origin points with a uuiforrn distribution.
The adopted value for ground flash density Ng was 3.5
flashes per km2/year, corresponding to La Plata region, It was useful to analyse present lightning performance of
obtained from reference [7]. refineries. This tool is based on the electrogeometric model

287
of the lightning process. According to this model the contain flammable liquids or gas under pressure
striking distance of a lightning stroke is expressed as a normally do not require lightoing protection, siuce such
function of the stroke current, as it is given by the equipment is well shielded from electrical strikes.
following most frequently accepted expression: Equipment of this type is normally well grounded and is
thick enough no to be punctured by a direct strike. They
(8) can be considered as objects being self-protecting.
Where: metallic tanks that had not been maintained in good
R: striking distance in metres conditions cannot be considered as self-protecting
I: stroke current in kA objects. Holes over the roof can be responsible for
k, n: empirical constants flammable concentrations of vapour or gas that can
result in a fire or explosion as a consequence of a
4.2 Data from installations of the refineries lightoing direct stroke.
metallic tanks used for storage flanunable substances at
Significant data research was performed in each atmospheric pressure, not necessarily have thickness
refinery. This task turned very large and heavy, as a enough to withstand a direct strike without being
consequence of data being not available, and: because punctured. Hence they were considered as objects to be
of many difficulties faced especially in some ryfineries. protected in simulations with realistic hypothesis.
Characteristics of refinery installations ~ch as
dimensions, height, construction materials, thickness, 4.4 Simulation of different cases
location, contents, and so on, were relevan' for the
studies. In addition, visual inspection accomplished In order to study refinery lightning performance different
during several visits to the installations c6mpleted conditions were simulated considering, or not, self-
missing data. t protecting behaviour of certain iustallations, and
Once the collection of data was made, eachl element considering, or not, presence of existing lightoing rods.
was represented, for simulations with BpNSUB Different hypothesis were assumed, consequently different
program, as a parallelepiped with four €artesian cases were analysed. The "most pessimistic" hypothesis is
coordinates and height over soil level (Figure 1). the one that ignores any existing lightning rods and any self-
protecting object. Then all probable combinations were ,
4.3 Protection characteristics made.
;
BLINSUB program requires a categorising of *1J of the Large number of lightning strikes was simulated falling over
elements being represented. This categorising :is based every refinery represented, for each determined case. The
upon its characteristics facing a lightning strike. proper number of lightoing strikes was calculated choosing a
Elements had to be categorised as objects btiing se!f- sufficiently long period of time, and by means of the average
protecting or as objects to be protected. i ground flash density Ng.
For lightning stroke current amplitude, the program uses a
if . . . i ' 1 i [
!!I ..-........L.·-·-·-..i.-·-...... ..1 ...........1! ·······-·-L....... ··+·· ···-·-}·-·-·-·-+· . . . . i.. ··! statistical distribution curve based on empirical data.
. j . ! i . . ' .
·············+·-···-·--1··············+-- ····-····+- ··-·--!- ······-··'1-··· ····---+.. -·-·-·-·l········..i····f·····-·-·- The adopted values of k and n constants of equation (8) were
. .......... .L......... ) ...•.•..... 4..,............j............... f............i. .............L ............. L......L.J............. . 10 and 0.65 respectively.

:=Ff~**~4~= 1 4.5 Simulation to analyse protection levels

Tilis program was also useful to analyse protection levels. It


reproduces the rolling of a sphere over the contour of each
··············i·--·-·---·.i-.... -....-~............. ~..............j ............'!-... ····---~............-.;........l... +.......-.... element represented in the simulated area To perform this
. -· -- r..............
...............•.:.........
. !
i' . . . . . l~ ............!i ············1[i ·······-···t··-·-·---·.l
""· simulation, an external file with one defined value for
..............:..-····-·-t-·-·········1·····........t ..-·-·-·-i-·- ·······-f..•· ·····-t----·----!-........T. t·· current amplitude should be used. According to the
electrogeometric model, equation (8), the radius of the
sphere, which is correlated with the desired protection level,
defines the stroke current amplitude.
Figure 1: Elements representation from YPF La Plata refitjery tanks
area i Such kind of deterministic simulations were accomplished
Some of the following considerations were useful to for the three protection levels selected.
accomplish previous categorising: Unfortunately this program do not permit vary strike
metallic tanks, vessels, and process equipment that incidence angle, hence every flash simulated is right vertical.

288
' 1,
sphere method. Alternative designs are shown in Figures 3
·-··i. ....... ·-+- - ···········+j ·········+··· ... ····i; ············i ·············j--··-- ···-··!, ....
......... !. . and4.
.. -............,...... ·······i······
.;. !.
···-·j···· ..... f ·········-·+j · · · · · · -~· -··· ·······.r-··
~ ·····'.

LPS

....!............. iil
·····-·····"
TANK :
··•............l
.........._J_
...... ~1:n· .. __, ......... -.lI ............ ,:......... ...., .............,. ..._ ...................
i ; ............ .. ..
~ ~

• 1 \
i ~~ ! ~
.......:..;>.~.!!?.""!""'".::c:.~-1+~-~-r-····· ..,.. L. ............! ________ ·-.~ ···········,i---- LPS ..............
LPS

l ;

t . . .l......~~~.:........:1 I I
; l
POOL
: :
Figure 2: Output from the simulation in the YPF La Plata refinery :: ::
tanks area , . . . .0 ...... 1

Actually, structures being taller than the sphere radius


could be struck if they don't have lateral protection.
Figure 3: LPS overhead ground wires design for protection oftanks and
However, simulations performed with the BLINSUB ' pools
program, will not declare lateral strokes on them.

4.6 Simulation results


ll'S
The period being chosen for statistical simulations was
3.000 years. It became enough since changes in the •
.............
sequence of flashes produce no difference in results.
Results obtained from the simulations were given in
two ways. One of them, by means of a map with all the
. \.:.:•
l. ....

...•}.
elements being represented and marks indicating
lightning strokes to elements categorised as objects to
be protected. The program was set to omit showing Figure;4: LPS four single masts design for protection of tanks.
lightning strokes to earth in the map.
All exterru'1 LPS proposed were isolated from the space to
By way of illnstration, Figure 2 shows the output-map be protect$, in order to avoid the ignition of any flammable
that indicates flashes striking elements categorised as air-vapour! mixture in the tank or pool surroundings, as a
objects to be protected. Compare this map with Figure consequenee of great heat developed along the lightning
l, where you can see all the elements being channel.
represented.
ln order to verify the protected zone defined by the LPS
Another output of the program is a list of each flash proposed deterministic simulations with BLINSUB program
striking to objects to be protected and ignores either were perfohned (rolling sphere method).
lightning strokes to self-protecting elements and
lightning strokes to earth. The list indicates for each
5 CONCLUSIONS
stroke: origin flash coordinates, stroke current
amplitude, and the element being struck.
At the moment, great investments in industrial areas
such as refineries in order to reduce lightning damage
4.7 LPS proposed designs
are consequence of lack of lightning considerations
during planning and designing periods in the past As a
Some LPS designs were proposed for either open air
result concern in prevention against lightning is
pools emanating flammable vapour, and metallic tanks
increasing nowadays.
requiring lightning protection, as they appeared to be
the most dangerous structures of refinery installations. Bad or poor maintenance especially in tanks and
process plant equipment is another possible cause of
Proposed designs consist in two basic types: one
future losses associated with lightning strokes.
performed with overhead ground wires and the another
with four single masts. All of these designs were An urgent solution is needed to prevent risky
calculated for a protection level I, applying the rolling consequence in open-air pools containing flammable

289
products. A different pool design, inherently self- Elecci6n de los niveles de protecci6n contra el rayo" IRAM 1997.
[3] !EC 1024-1: "Prot>:ction of structures against lightning. Part 1: General
protecting, is recommendable for future
principles", IEC Standards 1990.
installations. [4] IEC 1024-1-1: "Protection of structures against lightning. Part 1:
BLINSUB program resulted an acceptable tool to General principles. Section 1: Guide A- Selection of protection levels
for lightning protection systems", IEC Standards 1993.
detennine the frequency of lightning flashes to
[5] !EC 1024-1-2: "Protection of structnres against lightning. Part 1-2:
each structure located in the refinery being General principles. Guide B - Desi'gn. installation, maintenance arid
represented in the simulation. It also allows inspection of lightning protection systems", IEC Standards 1998.
reproducing a sphere rolling over all exposed [6] NFPA 780: "Standard for the Inst8!Iation of Lightning Prot>:ctiO.
surfaces. System", NFPA Standards 1995.
[7] Tres afios de registro con contadores de descargas atmosfericas en la
Pcia de Buenos Aires, Revista Electrotecnica, Diciembre 1975. ,
6 REFERENCES [8] API Recommended Practice 2003 "Protection Against Ignitiohs
Arising out of Static, Lightning, and Stray Currents-- Fire and Safety
[1] IRAM 2184-1: "Protecci6n de las estructuras contra las descargas Coordination". Fourth Edition, March 1982.
electricas atmosfericas. Parte 1: Principios generates. IR.Ai.VI [9] Peter Hasse, ..Protecci6n contra sobretensiones de instalaciones de baja
1996''. tensi6n''. ed. Paraninfo S.A, Madrid 1991.
[2] IRAM 2184-1: "Protecci6n de las estructuras contra las descargas
el6ctricas atmosfericas. Parte 1: Principios generales,_Secci6n 1:

290

Das könnte Ihnen auch gefallen