EN 50536:2011

SLOVENSKI STANDARD
01-julij-2011
=DãþLWDSUHGGHORYDQMHPVWUHOH1DSUDYH]D]D]QDYDQMHQHYLKW
Protection against lightning - Thunderstorm detection devices
Blitzschutz - Systeme zur Gewittererkennung
Protection contre la foudre - Dispositif de détection d'orage
Ta slovenski standard je istoveten z: EN 50536:2011
ICS:
07.060 Geologija. Meteorologija. Geology. Meteorology.
Hidrologija Hydrology
91.120.40 =DãþLWDSUHGVWUHOR Lightning protection
2003-01.Slovenski inštitut za standardizacijo. Razmnoževanje celote ali delov tega standarda ni dovoljeno.

EUROPEAN STANDARD
EN 50536
NORME EUROPÉENNE
May 2011
EUROPÄISCHE NORM
ICS 07.060
English version
Protection against lightning -
Thunderstorm warning systems
Protection contre la foudre -  Blitzschutz -
Dispositif de détection d'orage Gewitterwarnsysteme

This European Standard was approved by CENELEC on 2011-02-14. CENELEC members are bound to comply
with the CEN/CENELEC Internal Regulations which stipulate the conditions for giving this European Standard
the status of a national standard without any alteration.

Up-to-date lists and bibliographical references concerning such national standards may be obtained on
application to the Central Secretariat or to any CENELEC member.

This European Standard exists in three official versions (English, French, German). A version in any other
language made by translation under the responsibility of a CENELEC member into its own language and notified
to the Central Secretariat has the same status as the official versions.

CENELEC members are the national electrotechnical committees of Austria, Belgium, Bulgaria, Croatia, Cyprus,
the Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Iceland, Ireland, Italy,
Latvia, Lithuania, Luxembourg, Malta, the Netherlands, Norway, Poland, Portugal, Romania, Slovakia, Slovenia,
Spain, Sweden, Switzerland and the United Kingdom.

CENELEC
European Committee for Electrotechnical Standardization
Comité Européen de Normalisation Electrotechnique
Europäisches Komitee für Elektrotechnische Normung

Management Centre: Avenue Marnix 17, B - 1000 Brussels

© 2011 CENELEC - All rights of exploitation in any form and by any means reserved worldwide for CENELEC members.
Ref. No. EN 50536:2011 E
Foreword
This European Standard was prepared by the Technical Committee CENELEC TC 81X, Lightning
protection.
The text of the draft was submitted to the formal vote and was approved by CENELEC as EN 50536
on 2011-02-14.
Attention is drawn to the possibility that some of the elements of this document may be the subject of
patent rights. CEN and CENELEC shall not be held responsible for identifying any or all such patent
rights.
The following dates are proposed:
– latest date by which the amendment has to be implemented

at national level by publication of an identical

national standard or by endorsement
(dop) 2012-02-14
– latest date by which the national standards conflicting

with the amendment have to be withdrawn
(dow) 2014-02-14
– 3 – EN 50536:2011
Contents
Introduction . 6
1 General . 7
1.1 Object . 7
1.2 Scope . 7
2 Normative references . 8
3 Terms and definitions . 8
4 Thunderstorm phases and detectable phenomena for alarming .11
4.1 Introduction .11
4.2 Phase 1 – Initial phase (Cumulus stage) .11
4.3 Phase 2 – Growth phase .12
4.4 Phase 3 – Mature phase .12
4.5 Phase 4 – Dissipation phase .12
5 Classification of thunderstorm detection devices and their properties .12
6 Alarm method .14
6.1 General .14
6.2 Areas .14
6.3 Alarm triggering .15
6.4 Alarm information delivery .17
7 Installation and maintenance .17
8 Alarm evaluation .17
8.1 General .17
8.2 Evaluation of systems by using lightning location data .19
8.3 Fine tuning of TWS by processing archived data .19
9 Thunderstorms Warning Systems application guide .20
9.1 General .20
9.2 Procedure .20
Annex A (informative) Overview of the lightning phenomena .23
A.1 Origin of thunderclouds and electrification .23
A.2 Lightning phenomena .24
A.3 Electrical thunderstorm and lightning characteristics useful for prevention .25
Annex B (informative) Thunderstorm detection techniques .27
B.1 Introduction .27
B.2 Detection techniques and parameters to qualify a sensor .27
B.3 Location techniques .28
B.4 Thunderstorm detectors evaluation .30
B.5 Choosing a thunderstorm detection system .30
Annex C (informative) Thunderstorms Warning Systems application examples .31
C.1 Example n° 1 – TELECOMUNICATION TOWER .31
C.2 Example n° 2 – GOLF COURSE .33
C.3 Example n° 3 – WIND TURBINE FARM (including its maintenance) .35
Annex D (informative) Catalogue of possible recommended preventive actions to be taken .38
Annex E (informative) Example of TWS evaluation on a wind turbine site .41
Bibliography .43

Figures
Figure 1 ― Examples of different target shapes . 14
Figure 2 ― Example of the distribution of the coverage area (CA), the monitoring area (MA)
and the target area . 15
Figure 3 ― Example of an alarm. a) Locations of the lightning related events (LRE) in the defined
areas (coverage area CA, monitoring area MA, surrounding area SA, and target ); b)
temporal occurrence of the lightning related events (LRE); and c) timing of the alarm
according to the occurrence of the lightning related events (LRE) in the defined areas. Note:
surrounding area used in this figure is defined in 8.2) . 16
Figure 4 ― Introduction of the surrounding area (SA) for evaluation purposes . 19
Figure A.1 ― Adapted from Krehbiel (1986) . 23
Figure A.2 ― Standard lightning classifications . 24
Figure D.1 ― Possible preventive steps . 40
Figure E.1 ― CG lightning activity around the wind turbine for a period of eight years (a total of
2 480 strokes were reported) . 41
Tables
Table 1 ― Lightning detector properties . 13
Table 2 ― Contingency table . 18
Table 3 ― Identification of hazardous situations. 21
Table 4 ― Loss concerning people . 21
Table 5 ― Loss concerning goods . 21
Table 6 ― Loss concerning services . 22
Table 7 ― Loss concerning environment . 22
Table 8 ― Risk control . 22
Table C.1 ― Identification of hazardous situations . 31
Table C.2 ― Loss concerning people . 32
Table C.3 ― Loss concerning goods . 32
Table C.4 ― Loss concerning services . 32
Table C.5 ― Loss concerning environment . 32
Table C.6 ― Risk control . 33
Table C.7 ― Identification of hazardous situations . 33
Table C.8 ― Loss concerning people . 34
Table C.9 ― Loss concerning goods . 34
Table C.10 ― Loss concerning services . 34
Table C.11 ― Loss concerning environment . 34
Table C.12 ― Risk control. 35
Table C.13 ― Identification of hazardous situations . 35
Table C.14 ― Loss concerning people . 36
Table C.15 ― Loss concerning goods . 36
Table C.16 ― Loss concerning services . 36
Table C.17 ― Loss concerning environment . 36
Table C.18 ― Risk control. 37
Table D.1 ― Possible preventive steps . 39

– 5 – EN 50536:2011
Table E.1 ― Results of TWS evaluation based on archived lightning date for an 8-year period
(2000 to 2007), when some of the key parameters (size of MA, trigger parameters and dwell
time) were varied . 42

Introduction
Natural atmospheric electric activity and in particular cloud-to-ground lightning poses a serious threat
to living beings and property.
Every year severe injuries and even deaths of humans are caused as a direct or indirect result of
lightning:
– sport, cultural and political events attracting large concentrations of people may have to be
suspended and evacuated in the case of a risk of thunderstorm;
– power outages and unplanned interruptions of production processes;
– the wider use of electrical components that are sensitive to the effects of lightning (in industry,
transportation and communication) has led to a steady increase in the number of accidents per
year. In order to reduce this number of accidents and important material losses, it may be
necessary in some circumstances, to disconnect certain equipment from any incoming
installations;
– thunderstorms could interrupt all kinds of traffic (people, energy, information, etc.);
– activities with an environmental risk, for example: handling of sensitive, inflammable, explosive or
chemical products.
Lightning is also one of the causes of fires.
During the last decades, technical systems and systems devoted to real-time monitoring of natural
atmospheric electric activity and lightning have experienced an extraordinary development. These
systems can provide high quality and valuable information in real-time of the thunderstorm occurrence,
making it possible to achieve information which can be extremely valuable if coordinated with a
detailed plan of action.
Although this information allows the user to adopt anticipated temporary preventive measures, it
should be noted, however, that all the measures to be taken based on monitoring information are the
responsibility of the system user according to the relevant regulations. The effectiveness will depend
largely on the risk situation involved and the planned decisions to be taken. This document shows a
list of possible actions that is, however, merely of an informative nature.
It should be pointed out that lightning and thunderstorms, as any natural phenomenon, are subject to
statistical uncertainty. This means that it is not possible to achieve 100 % precise information on when
and where lightning will strike.
Standards dealing with lightning protection methods to limit lightning damages already exist. They do
not cover other potentially dangerous situations related to thunderstorms and lightning, that can be
dynamically prevented or reduced by temporary measures whose origin is a preventive alert provided
by a detection system.
– 7 – EN 50536:2011
1 General
1.1 Object
This European Standard provides information on the characteristics of thunderstorm warning systems
and information for the evaluation of the usefulness of lightning real time data and/or storm
electrification data in order to implement lightning hazard preventive measures.
1.2 Scope
This European Standard provides the basic requirements of sensors and networks collecting accurate
data of the relevant parameters informing in real-time about lightning tracking and range. It describes
the application of the data collected by these sensors and networks in the form of warnings and
historical data.
This European Standard applies to the use of information from thunderstorm warning systems (which
are systems or equipment which provide real-time information) on atmospheric electrical activity in
order to monitor for preventive means.
The scope of this document is providing:
– a general description of the available lightning and storm electrification hazard warning systems;
– a classification of thunderstorm detection devices and properties;
– guidelines for alarming methods;
– a procedure to determine the thunderstorm information usefulness;
– some examples of possible preventive actions (only for information).
A non-exhaustive list of activities to which this European Standard might apply is given below:
– people in open areas: maintenance people, labour, sports or other open-air activities,
competitions, crowded events, agricultural activities, farms and fisheries;
– wind farms, larger solar power systems, power lines, etc.;
– occupational health and safety prevention;
– safeguard sensitive equipment: computer systems, electric or electronic systems, emergency
systems, alarms and safety;
– prevention of losses in operations and industrial processes;
– prevention of serious accidents involving dangerous substances (e.g. flammable, radioactive,
toxic, and explosive);
– prevention in determined environments or activities with special danger of electrostatic
discharges (e.g. space and flight vehicle operations);
– operations in which the continuity of the basic services is needed to be guaranteed (e.g.
telecommunications, the generation, transport and distribution of energy, sanitary services and
emergency services);
– infrastructures: ports, airports, railroads, motorways and cableways;
– civil defence of the environment: forest fires, land slide and floods;
– managing traffic (e.g. airplanes) or wide networks (e.g. power lines, telecommunication lines) may
also benefits from having early detection of thunderstorms.

The following enumerated aspects are outside of this European Standard:
a) lightning protection systems. Such systems are covered by EN 62305 standards series;
b) other thunderstorm related phenomena such as rain, hail, wind, etc.;
c) satellite and radar thunderstorm detection techniques.
2 Normative references
The following referenced documents are indispensable for the application of this document. For dated
references, only the edition cited applies. For undated references, the latest edition of the referenced
document (including any amendments) applies.
EN 62305 series, Protection against lightning (IEC 62305 series)
3 Terms and definitions
For the purposes of this document, the following terms and definitions apply.
3.1
alarm
information indicating that the target is potentially subject of being affected by thunderstorms and the
accompanying lightning related events
3.2
cloud flash
lightning flash that never reaches the ground
NOTE 1 It can be an intra-cloud, a cloud-to-cloud or a cloud-to-air flash.
NOTE 2 By extension the term “intra-cloud” (IC) lightning sometimes encompasses the whole cloud flash family.
3.3
lightning flash to earth
CG flash
electrical discharge of atmospheric origin between cloud and earth consisting of one or more strokes
[EN 62305-1:2011]
3.4
coverage area
CA
area where a given warning equipment has a sufficient detection efficiency and/or accuracy to
elaborate a warning
3.5
detection efficiency
DE
percentage of actual lightning discharges that are detected and located by a sensor or a network
NOTE As cloud to ground flashes are often composed of several strokes there is a difference between flash
detection efficiency (DEf) and stroke detection efficiency (DEs). A flash is reported (detected) if at least one stroke
(first or subsequent) is detected and therefore DE is always equal or higher than DE .
f s
3.6
dwell time
DT
time that an alarm is sustained after all warning criteria are no longer met

– 9 – EN 50536:2011
3.7
effective alarm
EA
alarm where a lightning related event occurs in the surrounding area during the total alarm duration
3.8
excessive alarm duration
EAD
time after the last lightning related event occurred in the target waiting for the alarm to be released
3.9
failure to warn
FTW
occurrence of a lightning related event in the target for which no alarm was raised
3.10
failure to warn ratio
FTWR
ratio of failure to warn with respect to the total number of situations with lightning related events in
target
3.11
false alarm
FA
alarm never followed by lightning related events in neither the target nor the surrounding area
3.12
false alarm ratio
FAR
ratio of false alarms with respect to the total number of alarms
NOTE False alarm ratio is also known as false alarm rate.
3.13
field strength meter
FSM
device for continuous monitoring of the atmospheric electrostatic field associated with thunderstorms
(e.g. field mill)
3.14
intra cloud flash
IC
see cloud flash
3.15
lead time
LT
time between the start of an alarm and the effective occurrence of the first lightning related event in
the target
3.16
lightning flash
electrical discharge produced by a thunderstorm
NOTE This discharge may occur within or between clouds, between the cloud and air, between a cloud and the
ground or between the ground and a cloud.
3.17
dangerous event
LRE
lightning flash to or near the structure to be protected, or to or near a line connected to the structure to
be protected that may cause damage
[EN 62305-2:2006]
3.18
lightning stroke
single electrical discharge in a lightning flash to earth
[EN 62305-1:2011]
3.19
location accuracy
LA
statistical measure of the position difference between the actual strike point and the estimated location
NOTE Typically given as a median (50 %) location error.
3.20
monitoring area
MA
geographic area where the lightning activity or other parameters associated with the thunderstorms is
monitored in order to elaborate a warning valid for the target
3.21
physical damage
damage to a structure (or to its contents) or to a service due to mechanical, thermal, chemical or
explosive effects of lightning
3.22
preventive actions
actions of a temporary nature, taken on the basis of the preventive information and framed within the
emergency plans of each activity, service or collective
3.23
relevant alarm duration
RAD
time between the occurrence of the first and last lightning related event in the surrounding area while
the alarm was raised
3.24
return stroke
see lightning stroke
3.25
point of strike
point where a lightning flash strikes the earth or protruding objects (e.g. structure, lightning protection
system, line, tree, etc.)
[EN 62305-1:2011]
NOTE A lightning flash may have more than one point of strike.

– 11 – EN 50536:2011
3.26
surrounding area
SA
geographic area that surrounds and includes the target
NOTE Any lightning related event occurring in the surrounding area is potentially dangerous. This area is used
when evaluating a thunderstorm warning system to determine the false alarm ratio and other performance
parameters.
3.27
target
geographic area where a warning is needed in order to facilitate decision making and to activate
preventive actions before a lightning related event occurs in that area
3.28
thunderstorm
local storm produced by a cumulonimbus cloud and accompanied by lightning and thunder
3.29
thunderstorm detectors
equipment capable of evaluating one or more parameters associated with the electric mechanism of
the thunderstorm
NOTE Thunderstorm detectors may consist of a single detector or of a network of connected detectors.
3.30
thunderstorm warning system
TWS
system composed by thunderstorm detectors able to monitor the thunderstorm activity in the
monitoring area and some ways of processing to elaborate a valid warning related to the lightning
related events for a defined target
3.31
total alarm duration
TAD
time between triggering and the release of an alarm
3.32
warning
see alarm
3.33
warning level
current status of the alarm
4 Thunderstorm phases and detectable phenomena for alarming
4.1 Introduction
Four distinct stages can be identified during the thunderstorm life time cycle regarding detectable
phenomena: the initial phase, the growth phase, the mature phase and the dissipation phase.
4.2 Phase 1 – Initial phase (Cumulus stage)
Phase of cloud electrification by means of electrical charge separation within the cloud. The charges
are distributed in regions within the cloud and produce a measurable electrostatic field at ground level.
It is considered the first detectable phenomenon precursory of a thunderstorm.
NOTE Electrostatic fields may produce potential dangers such as electrostatic discharges (ESD) even in case of no lightning
activity.
4.3 Phase 2 – Growth phase
This phase, sometimes also called development phase, is characterized by the occurrence of first
intra-cloud IC (or cloud-to-ground CG) lightning activity. The first intra-cloud (IC) flashes appear after
certain development of the charge regions in the cloud. However in some situations there is no clear
time delay between the first IC flash and the first CG flash.
NOTE IC flashes typically represent the majority of the total lightning activity generated by a thunderstorm. Significant
variation in the IC/CG rate is observed for individual storms.
4.4 Phase 3 – Mature phase
This stage is characterized by the presence of both CG and IC flashes.
4.5 Phase 4 – Dissipation phase
This phase is characterized by the decaying of both IC and CG flash rates and the reduction of the
electrostatic field to the fair weather level.
5 Classification of thunderstorm detection devices and their properties
Thunderstorm detectors are classified in relation with the detectable thunderstorm phases depending
on the detectable phenomena. However, a thunderstorm detector can detect one or several
phenomena.
There are several ways to look at the means to detect thunderstorms in general and lightning strikes in
particular. One way is to look at the phase of the thunderstorm for which a detector is meant in
particular. Another way is to look at the frequency range of the signal emitted by a lightning strike that
is used by a sensor. A third way is to look at techniques that a sensor uses to detect a lightning strike
and to calculate its position.
For the phases of a thunderstorm the following phases are recognized, as explained in Clause 4:
– phase 1: initial phase;
– phase 2: growth phase;
– phase 3: mature phase;
– phase 4: dissipation phase.
For the classification of thunderstorm or lighting strike detectors the following classes are defined:
– class I: detectors of class I detect a thunderstorm over its entire lifecycle (phases 1 to 4);
– class II: detectors of class II detect IC and CG flashes (phases 2 to 4);
– class III: detectors of class III detect CG flashes only (phases 3 and 4);
– class IV: detectors of class IV detect CG flashes (phase 3) and other electromagnetic sources
with very limited efficiency.
The classes are explained in more detail in B.1.

– 13 – EN 50536:2011
The frequency ranges that are used in lightning detection are:
– DC: static and quasi static electrical fields;
– VLF: very low frequencies (3 kHz - 30 kHz);
– LF: low frequencies (30 kHz - 300 kHz);
– VHF: very high frequencies (30 MHz - 300 MHz).
All these phenomena to be measured result in different sensor and location techniques. Those
techniques may be distinguished as follows:
– MDF: magnetic direction finder;
– TOA: time of arrival;
– RFI: interferometry;
– FSM: field strength measurements;
– RF: radio frequency signal strength measurements.
NOTE This list is not exhaustive.
These detection techniques are described in some detail in B.2.
Table 1 shows the connection between the frequency range in which a detector may operate and the
phases, classes and typical ranges of operation for those detectors.
Table 1 ― Lightning detector properties
Technique Physical detectable Frequency Phase/s Main Secondary Typical Application
phenomenon class class sensor
range
km
FSM Electrification process DC 1, 2, 3, 4 I 20 Short range early
warning systems
MDF Electrical charges VLF 2, 3 III II No limit Low detection
motion efficiency and
location accuracy
– very long range
detection
MDF, TOA Electromagnetic LF 2, 3 III II 600 - 900 Long range –
radiation high location
(lightning current) accuracy for CG
detection
TOA Breakdown and VHF 2,3 II III 200 Medium range –
leader processes high location
accuracy for both
(IC/CG)
CG and IC
RFI Breakdown and VHF 2,3 II III 300 Medium range –
leader processes high location
accuracy for both
(IC/CG)
CG and IC
RF Electromagnetic LF 3 IV 100 Meteorological
radiation interest
(lightning current)
NOTE The main class is the class for which the detector is designed. The secondary class is the class or are the classes for
which the sensor is appropriate too.

More information on the properties and guidance in choosing a sensor for a certain purpose is given in
Annex B.
6 Alarm method
6.1 General
In order to let the user take all possible preventive actions, a thunderstorm warning system (TWS)
shall provide an alarm for a target where the lightning related event (LRE) represents a threat.
The identification of those LRE is deduced from the description of dangerous situations provided in
Clause 9. An alarm derives from monitoring the lightning activity, either or both CG and IC but also
other parameters such as the electrostatic field in the monitoring area (MA). Combinations with
additional meteorological observations are usually employed (e.g. meteorological radar). For detection
systems able to provide mapping information (lightning detection networks, radars, etc) it is possible to
track potentially dangerous thunderstorm cells thus improving the performance of TWS. Information
about TWS is described in Annex B.
The setup of an alarm includes three steps:
– areas definitions;
– alarm triggering criteria;
– alarm information delivery.
All three steps should be documented. Guidelines to set up an alarm are presented in this section and
some examples are included in Annex E.
6.2 Areas
6.2.1 Target area: precise description of the area which should include the physical extension where
the warning is needed. The target area can be limited to a single point (Figure 1a)), e.g. tower on
which workers are operating, limited size factory or can be extended (e.g. large buildings, wind farms,
golf courses: Figure 1b)). It is however recommended to use larger areas for safety reasons. In many
cases, it may appear simpler to limit the LRE to the occurrence of CG flashes and therefore adapt the
size and shape of the target in order to take into account all possible induced effects. For example, a
system sensitive to overvoltages on the power line can be warned for the occurrence of CG flashes in
a target encompassing the site but also the power line and its vicinity (Figure 1c)). Therefore, each CG
flash occurring in this target will be treated as a LRE able to cause the dreaded overvoltage. Thus, the
target also depends on the type of LRE and the effects that it could cause (see Clause 7).

a) Single point b) Arbitrary shape c) Including services
Figure 1 ― Examples of different target shapes
6.2.2 Monitoring area (MA): The size and the shape of the monitoring area should be adjusted
according to the type of the TWS (see Annex B), its capabilities (see Annex B, e.g. detection efficiency
and location accuracy), the shape of the target, the objectives and the performance of the alarm
system.
– 15 – EN 50536:2011
6.2.3 Coverage area (CA): Once the MA is defined, the detection system should have a CA that
includes the MA. When the CA does not cover the whole MA necessary to elaborate a reliable warning
on the target, it will be essential to juxtapose several elementary systems. The detection efficiency
(DE) and/or the location accuracy (LA) of the detection system within the range of the MA should be
known and their influence to the alarm performance should be considered.
CA
MA
Target
Figure 2 ― Example of the distribution of the coverage area (CA),
the monitoring area (MA) and the target area
6.3 Alarm triggering
In general, an alarm is triggered when the monitored information provided by the TWS is detected
within the MA. The criteria of triggering should be defined and depends on the characteristics of the
TWS and its performance within the MA (e.g. one or several CG flashes, one or several IC flashes,
certain electrostatic field level, electrostatic field polarity, and combinations of some criteria).
An example of a timing of an alarm is displayed in Figure 3.

NOTE Surrounding area used in this figure is defined in 8.2.
Figure 3 ― Example of an alarm. a) Locations of the lightning related events (LRE) in the
defined areas (coverage area CA, monitoring area MA, surrounding area SA, and target ); b)
temporal occurrence of the lightning related events (LRE); and c) timing of the alarm according
to the occurrence of the lightning related events (LRE) in the defined areas.
The lead time (LT) is the time available to conduct the preventive actions before the first LRE in the
target area may occur.
– 17 – EN 50536:2011
In order to avoid switching the warning level permanently, the lightning warning system shall use a
dwell time (DT) to sustain the alarm even if the alarm criteria are not met any longer. If the value set
for the dwell time is too large, the excessive alarm duration will rise significantly thus making the alarm
more costly (depends on the application). Note that systems able to accurately detect the end of an
alarm by any other means than the occurrence of lightning flashes in the monitoring area, such as for
example Class I (field strength meter FSM) systems, may not use the dwell time to release the alarm
but the occurrence of this end-of-alarm condition.
The total alarm duration corresponds to the interval between the alarm trigger to the end of the dwell
time (DT).
6.4 Alarm information delivery
A clear alarm delivery procedure and protocol should be defined to ensure that the alarm information
will be properly received by the end user.
It is mandatory to monitor faults of the thunderstorm detectors and communication links and notify the
end users of all possible detected faults that may affect the availability and the quality of the alarm.
7 Installation and maintenance
Any thunderstorm detectors shall be installed according to the manufacturer’s instructions and in the
best conditions for ensuring the fewest disruptions produced by its environment. For this purpose it is
highly recommendable to make a prior study of the proposed location in order to adapt the sensors of
the system to the specific conditions of the site.
The installation of thunderstorm detectors is prone to be affected by multiple factors, so, any new
installation may need a prior adjustment period before it is considered to be working at its optimum
level. This adjustment shall be made by the system’s manufacturer or by a technician specifically
authorized by this manufacturer.
Maintenance of the systems integrated in a TWS, including alarm delivery is indispensable. The
precision of the information provided by a TWS is directly determined by the physical conditions of its
sensors, their environment (i.e. growing vegetation, buildings, towers, etc.), communications links
between the sensors and the TWS as well as between TWS and end users. So, it is considered
necessary to carry out the maintenance tasks every year or even at shorter periods according to the
manufacturer’s recommendations.
All these installation and maintenance recommendations are really a key factor for successful warning
system.
8 Alarm evaluation
8.1 General
By evaluating the operation of the TWS, it is possible to optimize its parameters and then improve the
quality and the reliability. The alarm can thus be better adapted to the end user applications.
Performance evaluation results in extremely valuable information for future alarm settings, preventive
actions improvements, and increases the knowledge of the target lightning environment.
It is recommended to establish an evaluation procedure. In this procedure the user should provide
information about previous experiences (e.g. number of alarms, failure to warn, false alarms,
damages, etc.) during a particular alarm setup.

The evaluation can be performed in different ways depending on the availability of validation
information, such as:
– experience and good sense: climatology, local observations, unrealistic alarm durations, etc.;
– cross-correlation with other sources of information: data from other lightning location systems,
meteorological radar, satellite, etc.;
– processing archived data for systems that are able to record all the information useful for
elaborating warnings. This is the only way to fine tune and verify the settings of the alarm
parameters.
The main performance data of a specific TWS is:
– the false alarm ratio (FAR) determined as the ratio of the observed false alarms (FA) to the total
observed alarms (FA+EA);
FA
FAR= (1)
FA+EA
– the failure to warn ratio (FTWR) determined as the ratio of the number of failures to warn (FTW)
to the expected total number of alarms (FTW+EA);
FTW
FTWR= (2)
FTW +EA
– the distribution of lead time (LT);
– the distribution of excessive alarm duration (EAD).
Table 2 summarizes how effective alarms (EA), false alarms (FA) and failure to warn (FTW) are
counted.
Table 2 ― Contingency table
LRE did occur in the SA LRE did not occur in the
SA
Alarm was delivered EA FA
No alarm was delivered FTW -
The main parameters that can be adjusted to improve the performance of a TWS are:
– the alarm trigger criteria in the MA;
– the size and shape of the MA;
– the dwell time (DT).
A change in parameters will always lead to compromises, for example:
– increasing MA size will increase the number of alarms, the lead time (LT) but also the false alarm
ratio (FAR) and excessive alarm duration (EAD);
– reducing MA size will increase the failure to warn ratio (FTWR) but decrease of the false alarm
ratio (FAR) and lead time (LT);
– increasing the sensitivity of the triggering criteria will decrease the failure to warn ratio (FTWR)
but could increase the false alarm ratio (FAR);

– 19 – EN 50536:2011
– reducing the dwell time (DT) will reduce the excessive alarm duration (EAD) but also tend to
artificially increase the number of alarms and reduce the lead time (LT).
According to the warning applications, the goal of performance optimization can be different:
– a minimum false alarm ratio (FAR) and excessive alarm duration (EAD) is required in applications
where the cost of service interruption is huge;
– a minimum failure to warn ratio (FTWR) is required in applications where human safety is
involved;
– a sufficient lead time (LT) is required in application where preventive actions can be long to
activate.
8.2 Evaluation of systems by using lightning location data
Lightning location data is available from many sources (lightning detection networks, satellite
observations, etc) almost everywhere with different quality in terms of detection efficiency (DE) and
location accuracy (LA). This data can be used as a pseudo-ground truth to evaluate the performance
of the TWS keeping in mind the limitation due to the given detection efficiency (DE) and location
accuracy (LA). Indeed, a poor detection efficiency (DE) of the validation dataset will have a tendency
to artificially increase the false alarm ratio (FAR).
In the process of evaluating a TWS it is necessary to introduce a surrounding area (SA)
encompassing the target as shown in Figure 4 in order to confirm the efficiency of alarming. Indeed, in
the case the target is warned although it never sees any LRE, the occurrence of some LRE in the very
close neighbourhood of the target (as defined by the surrounding area) indicates that the risk is in any
case, high and this situation may not be treated as a false alarm (FA). On the other hand, a target
warned while no LRE has been recorded at all, clearly indicates a malfunction of the equipment and
should be treated as a false alarm (FA). Moreover, the introduction of the surrounding area (SA)
allows for taking into account the limited location accuracy (LA) of the validation dataset.
Coverage Area
MA
SA
Target
Figure 4 ― Introduction of the surrounding area (SA) for evaluation purposes
8.3 Fine tuning of TWS by processing archived data
Some TWS have the ability to save raw data (lightning locations, E-field, etc.) over a long period that
can be used to optimize the warning parameters. According to the targeted performance of TWS (low
failure to warn ratio, long lead time, etc.) it will be possible to check the sensitivity of desired metrics
when the warning parameters (size and shape of MA and tri
...