SURGE PROTECTION, WHY USE IT IF IT IS NOT REQUIRED BY LAW?
by G R MacKenzie, A V Schneider, Product Marketing Manager, Phoenix Contact (Pty) Ltd,
P O Box 916, Ferndale, 2160, South Africa: e-mail:
Product Manager, Lightning and Surge Voltage Protection, Phoenix Contact (Pty) Ltd,
P O Box 916, Ferndale, 2160, South Africa: e-mail:
Abstract
In January of 2001, the SABS issued a revision to the document SABS 0142-1 which required that surge protection devices “shall be fitted to all new electrical installations irrespective of the lightning ground flash density”. This created a violent reaction from the electrical community and eventually the wording was revised in SANS 10142-1:2003 to read that “Surge protection devices (SPDs) may be installed…”. Why did this happen? Are SPDs still required by law? What is the role of the SABS? Who decides if SPDs are required, and if they are, which SPDs, how many SPDs and what will they cost? It’s a little like life insurance, you know you should have some but are not sure how much to spend or if it will work when you need it.
Introduction
This has been a very topical subject over the last few years due to the sudden changes in legislation, huge press exposure and subsequent modifications. Certainly today there is no apparent legislation requiring the installation of any type of surge protection into electrical installations. This of course also does not mean that surge protection is not required! So who decides if it is required? Should it be the designer, the owner of the installation, or the user of the installation. If it is required, does the SABS provide any guidelines about the selection and installation of SPD’s.
In this paper we would like to explore the role of a standards authority like the SABS and specifically the details of the SABS/SANS documents. Ignoring for the moment the legal requirements of the SANS 10142-1:2003 we would like to question whether the documents provide value to the user of electrical equipment in South Africa, or Africa for that matter as we are ignoring legal implications, and if they do add value, how to use these documents.
The objective of sans10142-1 (briefly)
The SANS 10142-1:2003 is considered to be the overriding standard for the wiring of all low voltage electrical installations. This document where relevant refers to other local and international standards which should be used to supplement the text of this base standard.
From Page 3 of SANS 10142-1:2003 we read…
“The aim of this part of SANS 10142 (SABS 0142) is to ensure that
people, animals and property are protected from hazards that can arise
from the operation of an electrical installation under both normal and fault
conditions.”
The history
Before starting a discussion of the status today, it is important to look at the detail of what happened over the last few years.
In 2001 the then SABS 0142 document was revised to include the installation of surge protection devices as mandatory in all electrical installations. Section 6.7.6 of this document read…
“Surge protection
Surge protection devices (SPDs) shall be installed to protect an installation
against overvoltage surges such as those due to switching operations or
those induced by atmospheric discharges (lightning).”
There was an immediate reaction to the introduction of the compulsory requirements which provided some very interesting reading in the press over the next few months. In the short time provided today, lets jump to the conclusion – in January 2003 this section was very simply changed to read that the SPD’s “may” be installed and the rest of this section was moved into Annexure L which is defined as being “informative” only.
Amendment 1 in the table of changes states…
“ amend the requirements for
surge protection, and change them to
recommendations (included in annex L),”
Further in the foreward on page 2 is stated…
“Annexes B, C, D, E, F, G, J, K, L, M, N, O, P, Q and R are for information
only.”
Why did this happen?
I would like to in a very simple way explain my understanding of the logic behind this change. The arguments can be represented by two questions.
- The first is a simple discussion of cost. In very many low cost installations, for example low cost housing where pre-payment meters are installed, the required SPDs would typically cost more than the pre-payment meter itself.
Where a utility is installing such a pre-payment meter, is it required to install protection for this low cost device. Typically a pre-payment meter will have some kind of protection fitted and it is cheaper to replace the entire meter should surge damage be experienced.
The cost implication to utilities and municipalities of this kind of requirement for surge protection was simply too high.
Perhaps a more compelling argument however refers back to the objective of SANS 10142-1 in section 3 “…to ensure that people, animals and property are protected…”
- Does the installation of an SPD in any way ensure that people, animals and property are protected?
Well certainly not people or animals, that is not normally the function of a Class II SPD as was required by the then current version of SABS 0142-1. What about property? Yes – but the question was, whose property?
If the property is for example the pre-payment meter, it makes no sense to install a protection device as discussed above. If the owner of said property does not see value in the protection of that property due to the low cost, how can they be legally forced to do so?
If the owner of the property (electrical equipment) inside the dwelling, perhaps a TV or fridge, requires protection, it is still not the responsibility of the utility to provide that protection. It must surely be the responsibility of the owner of this property to provide the protection required, if they see it as being valuable.
Although there is perhaps a third discussion about the moral obligation to provide a safe electrical supply, this is a very subjective argument and there is no time in this paper to discuss it. Certainly however according to the fundamentals of SANS 10142-1, there is no legal requirement for SPD’s.
So the SANS document was amended. This document no longer tells us whether SPDs are required or not. What does it tell us?
The sans 10142-1:2003
Today this is a 357 page document. Throughout the document, references are made to lightning and surge voltage protection. Below we would like to summarise the requirements and information contained in SANS 10142-1 referring to this subject…
Section 2 of SANS 10142-1 contains the following text…
“The following standards contain provisions which, through reference in
this text, constitute provisions of this part of SANS 10142…”
Each of the standards below refer in some way to lightning and surge voltage protection and will be covered in some way.
“SANS 10313 (SABS 0313), The protection of structures against lightning.”
“SANS 61000-4-5/IEC 61000-4-5 (SABS IEC 61000-4-5), Electromagnetic
compatibility (EMC) – Part 4: Testing and measurement techniques –
Section 5: Surge immunity test.”
“SANS 61312-1/IEC 61312-1 (SABS IEC 61312-1), Protection against
lightning electromagnetic impulse – Part 1: General principles.”
“SANS 61312-2/IEC TS 61312-2 (SABS IEC TS 61312-2), Protection
against lightning electromagnetic impulse (LEMP) – Part 2: Shielding of
structures, bonding inside structures and earthing.”
“SANS 61312-4/IEC 61312-4 (SABS IEC 61312-4), Protection against
lightning electromagnetic impulse – Part 4: Protection of equipment in
existing structures.”
“SANS 61643-1/IEC 61643-1 (SABS IEC 61643-1), Surge protective
devices connected to low-voltage power distribution systems – Part 1:
Performance requirements and testing methods.”
“IEC/TS 61312-3, Protection against lightning electromagnetic impulse –
Part 3: Requirements of surge protective devices (SPDs)”
Section 4.3 reads…
“Applicable standards
Table 4.2 gives a list of commodities, the applicable safety
standards and recommended performance standards. The commodities
given in column 1 shall comply with the standards given in column 3 and
it is recommended as good practice to comply with the standards given in
column 4.”
(This is also referred to in the forward on page 2 where it is stated…
“Table 4.2 contains a list of the applicable standards for the components
that may be installed in an electrical installation.”)
Table 4.2 (continued) on page 64 shows…
1 / 2 / 3 / 4Commodity / Scope / Safety
standard / Recommended
performance
standard
… / … / … / …
Surge arresters for low-voltage systems / 1 000 V / SANS 61643-1
… / … / … / …
As mentioned already, but as part of the current standard we will repeat…
“Surge protection
Surge protection devices (SPDs) may be installed to protect an installation
against overvoltage surges such as those due to switching operations or
those induced by atmospheric discharges (lightning).
See annex L for the installation of SPDs”
Most relevant in this clause was that the word “shall” used previously was changed above to “may”.
Note that Annex L is informative only, although the wording may indicate otherwise, none of what is mentioned is required by SANS 10142-1. We will not discuss this annex in detail due to this fact and if included may confuse the standard required by SANS 10142-1. Some information dealing with subjects like installation, zones and impulse withstand categories are very useful in summarising information covered also in other standards and will be referred to later.
The interpretation?
SANS 10142-1:2003 seems therefore to really only say 2 things…
- SPDs may be installed to protect an installation, and
- If SPDs are installed, they should comply with the performance standard SANS 61643-1.
It seems the decision to install lightning and/or surge voltage protection is left entirely up to the owner of the equipment which may or may not need protecting.
The basics of lighting and surge voltage protection
Before looking further at why or how to protect against lightning or surge voltages, it is important to establish some ground rules. Lets look at the causes and effects of transient voltages and the technology required to protect against these transients, and how this is defined in SANS 61643-1.
(Note: This paper is not intended as a training seminar on lightning and surge voltage protection. Detailed training courses are available from various bodies in South Africa, the writers will be glad to forward such information on request. The information is given below as introduction only)
Types of transient voltages
There are considered to be four classes of transient voltages…
- Lightning Electromagnetic Pulse (LEMP)
- Switching Electromagnetic Pulse (SEMP)
- Electrostatic Discharge (ESD)
- Nuclear electromagnetic Pulse (NEMP)
As ESD typically will not effect a completely manufactured/enclosed device and the risk of experiencing NEMP is we hope not significant, we will focus only on LEMP and SEMP.
LEMP: LEMP is of course caused by the direct or indirect effect of a lightning strike on an electrical installation. LEMP will effect an electrical installation typically in one of two ways shown in Figure 2 below.
Galvanic coupling occurs when lightning directly strikes a building or installation, or strikes the externally mounted lightning protection system (LPS). A high impulse current (iimp) is galvanically coupled into the LPS. As the LPS is coupled to earth and the electrical distribution system of the installation is likewise coupled to earth, this iimp is split into i1 and i2. The current i2 is considered typically to be 50% of iimp and is coupled into the electrical network of the installation.
Figure 1: LEMP coupling to electrical installations
Inductive coupling of course results from the inductive effect of a large change of current flowing through the LPS. The size of this induced surge is more difficult to predict.
SANS 61312-1 (referred to in SANS 10142-1) defines the parameters of a lightning current as follows in Figure 2.
Figure 2: SANS 61312-1 Lightning current parameters
The protection level I, II, III or IV (or IEC risk category) can be determined also using SANS 61312-1, however it can be safely assumed that a 200kA lightning strike or category I occurs in <1% of lightning strikes.
SEMP: A switching electromagnetic pulse or SEMP is a very different type of transient. Figure 3 below shows typical causes of SEMP. A fuse blowing or large contactor opening can cause a voltage transient to be introduced into the electrical network.
This transient however by nature is much faster as it typically has much less energy. (There is no energy introduced into the network as is the case in a lightning strike) Whereas in Figure 2 above the typical lightning strike is considered to have a wave shape T1/ T2 of 10/350s, a switching pulse is considered to have a typical wave shape (used for testing of a Class II arrester according to SANS 61643-1) of 8/20s, more than 17 times shorter.
Figure 3: Typical causes of SEMPs
Protecting against transients
The amount of energy required to damage a diode or IC is typically <0.001Ws. A transient of 1000V with a half life (T2) of 20s would require a current of roughly 100mA to produce this amount of energy. It is also clear however that an SPD required to arrest 100mA in 20s (T2) is not the same as the SPD required to arrest an LEMP of 200kA. This has resulted in the definition by the IEC, and adopted by SABS, of different SPD classes, presented in SANS 61312-1.
These devices are classified as…
- Class I SPD = lightning current arrester
- Class II SPD = voltage clamping device
The technology used in these two types of SPDs is very different, however the SPDs always have the same purpose. As a transient travels along an electrical wire towards an electronic device, the transient voltage needs to be equalized to the earth potential of the electronic device to prevent damage to the electronics caused by a discharge of energy through the electronics. See Figure 4. The SPD’s are indicated by S1, S2 and S3 showing a short circuit connection between conductors and between each conductor and earth. This short circuit diverts current to earth for the duration of the LEMP or SEMP surge (typically not more than 1 or 2 ms) thus clamping the surge voltage, and then must restore an open circuit to allow normal operation of the electrical circuit. These devices may act independently or together depending on the nature of the transient.
SPD CLASS II: Switching electromagnetic pulses (SEMP) are typically controlled through the use of voltage clamping devices such as Metal Oxide Varistors (MOV’s) or Silicon Avalanche Diodes (SAD’s). MOV’s are able to handle larger discharge currents and SAD’s are typically faster but they do the same thing. As the surge voltage reaches a predetermined level known as the protection level, so the electrical characteristic of the MOV or SAD causes a current discharge to earth, so clamping the surge voltage and diverting the surge current from the protected electronics.
There is a trend today to connect multiple SAD’s to have a higher discharge capacity and keep the fast response time, this however becomes expensive. When one considers the 8/20s characteristic of SEMPs, the additional cost of an SAD with 5ns response versus an MOV with 25ns response against the 8,000ns rise time T1 seems hard to justify.
Figure 4: Functional principal of an SPD
SPD CLASS I: MOV or SAD technology is found however to be prohibitively expensive, and often not effective in the protection against LEMP impulse currents any higher than just 4 or 5kA using a 10/350s wave shape. In these cases it is found that a spark gap type device can provide more than 50kA per conductor protection against multiple lightning strikes without damage to the SPD or protected equipment.
Today spark gap technology has been advanced with the introduction of a “voltage detector unit”, typically including an MOV which can create a controlled ignition of the grounding spark. Figure 5 below shows the measured response of a circuit injected with a 15kA 10/350s wave shape pulse. This result shows that the maximum voltage umax is clamped to just 569V while iimp rises to 12,7kA in 10s and falls to <50% in 350s as required. As will be reinforced later, spark gap technology provides effective and low cost protection of the electrical network.
Figure 5: test result from Class I triggered spark gap arrester
with 15kA 10/350s wave shape impulse
Combined CLASS I and CLASS II SPDs: Note that the MOV in the triggered spark gap above is functioning as a voltage detector and is not used to discharge any LEMP induced current. It is possible to fit a Class II type MOV protection into the same housing as the Class I spark gap to provide a combined Class I and Class II arrester. Two warnings here; a) the MOV is an electronic device and will degrade over time and use. In a combined arrester, degradation or damage to the MOV will result in the necessary replacement of a perfectly functioning spark gap device, b) ensure that the specification of the combined device quotes that it has been approved for Class I and Class II according to SANS/IEC 61643-1 and includes both iimp (10/350s) and imax and in (8/20s) test values.
The risk
There appear to be two risks. First the risk of a lightning strike, second the risk of the effects of SEMP. But how does one define risk? Is it the calculation of the possibility of an electrical installation being struck by lightning, or is it the financial calculation of cost of investment versus cost of loss?
The risk of a lightning strike (lemp)
A very large amount of research and calculation has gone into this subject. The SABS here has helped with the standard document SANS 10313, (referred to in SANS 10142-1) which provides a risk calculation for…
The accepted annual frequency of lightning flashes: which takes into account the type of structure, the contents of the structure and the effect of consequential losses.
The expected frequency of direct lightning flashes to a structure: which looks at the average annual ground flash density, the collection area of a structure and an environmental coefficient depending on location.
The evaluation of a lightning protection system (LPS): which uses the above calculated values to determine the required LPS efficiency.
4.4.4 of this document then states that if the efficiency of the installed LPS is lower than the required efficiency, additional measures of protection shall be provided…for example…measures to mitigate the effects of lightning-induced overvoltages on sensitive equipment.
The risk of switching electromagnetic pulses (semp)
No mention is made here by SABS of how to calculate this risk. Surely there are too many variables, i) the quality of the power supply; ii) the number of installed electrical devices, fuses, contactors and breakers etc; iii) the quality and age of the installed devices; iv) the ongoing maintenance of the electrical system; and so on. Perhaps the only way to do this is to measure the transients on the network over a period and then hope that nothing changes?