Comparing UL and IEC Test Methods for Surge Protective Device(SPD)

Comparing UL and IEC Test Methods for Surge Protective Device(SPD)

Both UL and IEC go to some length to produce tests which will simulate various fault conditions an SPD may encounter during its operation, and then to evaluate that the device is able to either withstand or disconnect from these in a safe manner.

Some have argued that the UL standard is probably more thorough in the area of safety testing than its IEC counterpart which arguably has a greater emphasis on performance testing. If there is any truth in this statement, it may be as a result of the different environments which SPDs encounter between IEC and ANSI based countries. For example, the issue of “loose neutrals” is more common to the 120/240V 3W+G single phase supply used in North American countries5. The US is also particularly aware of the risks which fire poses to its residential dwellings which are predominantly wood construction, rather than brick and mortar as is common to European countries.

Weakness in UL test methods

The UL standard allows a manufacturer to adopt“containment” measures as a means to pass the various current tests described above (Section 39 in the standard).The only condition being that the usual pass criteria are met(e.g. tissue paper and cheese cloth must not burn and there must be no expulsion of molten material etc).

Weakness in UL test methods:The UL standard allows a manufacturer to adopt“containment” measures as a means to pass the various current tests described above (Section 39 in the standard).The only condition being that the usual pass criteria are met(e.g. tissue paper and cheese cloth must not burn and there must be no expulsion of molten material etc).

The problem with this is that there is no guarantee that the product will internally fail the same way in each case. By not requiring that a specific component be the current isolator (such as a fuse or thermal disconnect) is essentially allowing an uncontrolled behaviour. It is hard to argue with the logic that if the test were to be conducted at a different current, the uncontrolled internal failure would not violate(explode) the housing!

It is also troubling that certain measures UL require of manufacturers to ensure conformance in production, work unwittingly against the good intents of manufacturers to in corporate safe disconnect technologies into their products,and instead steer them down the arguably unsafe“containment” path.

For example, UL requires an SPD which includes a “non-recognised” disconnect to have this disconnect evaluated to a standard called UL 61691 (Ref. IEC 60691). This standard was designed to evaluate thermal-links, the small fuse-like components with set melting points which are utilised in a host of household electrical appliances to disconnect in the event of the temperature being exceeded.

The intention of this standard was never to evaluate the generally more robust thermo-mechanical disconnectors engineered into SPDs. As a result, the standard only has provision to evaluate the effective operation of such disconnects to a maximum short circuit current (SSC) of some hundreds of amps, while SPDs need to be rated with SCCRs equal or greater than that of the power system to which they will be connected – in most cases some tens of thousands of amperes.

Given this limitation – UL not only requires SPD manufacturers who incorporate such “non-recognised” disconnects in their products to test to this (inappropriate)UL 60691 standard, but also to pay for annual follow-uptesting at the required SCCR. Such testing typically runs inexcess of $50k – a large annual burden to any manufacturer.

If on the other hand, the manufacture does not include any sort of disconnect in his product, and simply relies on containment measures, it is not subject to any further testing provided the one sample passes.

There are elements here of UL trying to fit a square peg into a round hole (enforcing a SPD disconnects to go through an inappropriate standard), and not recognizing the excess burden they are placing on manufacturers (requiring annual follow up services) is steering SPD design to less than safe practice.

Weakness in IEC test methods

One area where criticism of the IEC 61643-1 document is probably justified, is in the method of determining (and declaring) the short-circuit current withstand rating Isc of an SPD.

The IEC test method involves the replacement of the“active” components of the SPD with copper blocks(dummies). This creates an artificial situation which it is argued does little more than test the disconnector (external or internal) and internal connections, rather than meeting the requirement that “an overstressed SPD shall withstand the power short-circuit currents that may occur in service.”

Furthermore, this test fails to evaluate one of the more acknowledged causes of SPD induced fire – that created when the active components catastrophically fail and in so doing, deposit semi-conductive metallization throughout the SPD, or cause internal conductive plasmas that can start follow-current arcing and burning.

It is important to understand that these active components are the main source of heat generation in the event of SPD failure (especially at intermediate current faults) and therefore the primary initiator of fires. Short-circuiting this component removes the potential heat source and leaves the IEC test method open to criticism.

It suffices to say that IEC SC37A/WG5, which is responsible for the development and maintenance of the standard IEC 61643-1, has established a task force to review the present test methods used to evaluate the disconnection of SPD’s at end-of-life and during fault conditions. The difficulty faced by this task force is how to devise a more appropriate test which will evaluate the safe disconnection of a fail SPD while not creating artificially abnormal conditions to induce this behaviour in the first place!