Surge Protection Devices (SPD's) - Shunt vs Series.

The purpose of a Surge Protection Device (SPD) is to limit the risks posed through transient over-voltages reaching sensitive mission critical equipment, that are commonly associated with:

• External sources, such as lightning, and
• Internal sources, such as switching transients

No SPD is perfect, in fact some let-through voltage over and above the nominal mains voltages will always reach the equipment, which is acceptable provided that the let-through voltage is within the tolerance envelope of the connected equipment.

Electronic and electrical equipment have varying levels of susceptibility to transients, where electrical equipment such as motors, HVAC, and lighting, are generally considered robust, and do not require such a high level of protection, where they can typically withstand transient impulses in the region of 600 to 1000 (Vpk) ( L-N).

However "semi conductor" based equipment such as Servers, Computers, PLC, PABX, Network Switch, SCADA, RTU, UPS’s, Battery Chargers, Rectifiers etc, are more sensitive to the peak voltage (Vpk), as well as the high rates of voltage rise (dV/dT) typical of lightning transients, and therefore require a much higher levels of protection.

In Australia there are generally two (2) types of SPD's.

• Primary - Shunt connected "Surge Diverters" which offer a "Coarse" level of protection.
• Secondary - Series connected "Surge Filters" which offer a "Fine" level of protection.

                    Fig 1. Shunt vs Series connection arrangements 

Simplified explanations as to the difference between "shunt" and "series" protection are as follows:

"Shunt connected" SPDs are typically located at the service " Point of Entry" (Main Switchboard) and are wired in "parallel" with all downstream connected equipment.

As this is a parallel method of connection, a portion of the transient energy is always shared with all downstream connected equipment, where the SPD is seen as a low impedance shunt path. These Surge Diverter's takes the main brunt of the energy away, and we say these offer a "Coarse" level of voltage protection.

"Series connected" SPD's are typically located "at" the sensitive equipment requiring protection, where any transient must travel through (not by) the SPD to get to the connected equipment.

This type of SPD includes a "low pass" (L-C) filter section which is current sensitive and must be sized to the load, and which is wired in series with connected load. This filter section acts to modify the pre-clamped waveform, which reduces the high rates of voltage rise (dV/dT) and current rise (dI/dT) from reaching the more voltage sensitive equipment.

Surge Filter's subsequently offer a very low let through voltage, and we say these devices offer a "Fine" level of protection.

The ANSI/IEEE C62.41 Cat B Waveform can be used as a standard test to determine the let through voltages for SPDs and can highlight the comparison of let through voltage for Series and Shunt connected SPD's when subjected to this waveform.

Fig 2. ANSI/IEEE C62.41 CAT B Open Circuit Waveform 6kV 1.2/50 us 3kA 8/20 us. Courtesy ERICO 

The typical results are shown below in Fig.3 where an ANSI/IEEE C62.41 Cat B Open Current has been impressed upon the connection terminals of both Shunt and Series connected SPD's.

Fig 3. Let though voltages for Series and shunt connected SPD's at Cat B. Courtesy ERICO

As can be seen with the Shunt connected SPD, there is no reduction in the impulse rise time resulting in a significant front end spike being passed through to the connected equipment, whereas the Series wired SPD presents a smoother voltage let through, that is less than half that of the shunt diverter.

In a previous life (late 80's/ early 90's) I worked for two (2) well known UPS manufacturers "Sola" and "Liebert", and I cannot recall how many times I had seen significant centralized UPS damaged during localized thunderstorms, but it was a lot!

I can also attest that our UPS service techs were ALWAYS at their busiest, following localized thunderstorms, yet many IT Managers, and Facilities Management professionals seemed content that their singular investment in UPS was sufficient to cater to power quality risks posed by lightning, despite the fact that this was hardly the case, and the reality was lightning posed a significant to their UPS, and they were under a false sense of security that UPS provided anything resembling an appropriate level of protection.

The fact of the matter was there was very little understanding as to the capability of UPS as being suitable for dealing with lightning transients, where the reality was/is that UPS were/ are just as susceptible to lightning transients, as are the critical loads they were tasked to protect.

The fact is that that uninterruptible power systems DO NOT include any suitable lightning protection components within their design or capabilities, other than the front end rectifier component states a max surge withstand of ANSI C62.41 Cat B, which means the rectifier component is limited to a total protection level of 6 kV, 3 KA (@ 8/20 us).

The reality is that a typical lightning transient will cut through this level of surge withstand, like a hot knife through butter, and I hold vivid memories of significant UPS damage resulting from lightning transients , which had included:

• Two (2) x 20 kVA UPS (from a Diamond mine) which had been directly affected by lightning current where there was not a single salvageable part on both systems.
• A 30 kVA UPS at a local Perth brewery which failed during a localized thunderstorm, after suffering from serious rectifier damage. I understand this unit never recovered and was later scrapped.
• A 60 kVA UPS at a local correctional facility which was significantly damaged during a localized thunderstorm, where an induced transient entered the equipment via the field cabling and which took out the back end inverter, requiring the unit to be scrapped.

What was interesting regarding this particular correctional facility incident was that the damage resulted from a lightning transient inducing upon the output field cabling's, and subsequently entering the UPS via the output inverter section.

Further this I recall that a particularly competent Electrical Engineer (Yes , that would be you DW) of the former Building Management Authority (BMA) had previously specified "reverse wired" Surge Reduction Filters to be placed on the output of UPS he specified at a nearby correctional facility, and application that acted similarly to feed uninterruptible power to the various external electrical cubicles via exposed field cabling's out to the sterile zone inside the perimeter fence.

This practice of specifying Surge Reduction Filters on the input and output of UPS stopped after the aforementioned Engineer accepted a management promotion outside of the design area, and so such common sense practice was lost.

Nowadays I still visit many sites where significant investment in UPS has been installed, and whilst we do see an increased use of SRF's on critical equipment UPS and surge filters on critical circuits, there is little understanding as to the suitability of the UPS to act as a lightning protection device, and such equipment is largely unprotected.

Additionally think of what happens when ;

• A long duration outage exhausted the battery autonomy
• During a fault condition, the static bypass will transfer the critical load to raw mains, (in order to protect itself with no consideration of the connected critical load)
• During maintenance, the critical load would always be connected to raw mains.

UPS are not inherently self protecting.

Lightningman are specialists in the supply and application of surge diverter and surge reduction filters, from small 3 amp din rail versions, up to 3 Phase 2000 amp per phase models.

Grant Kirkby.

www.lightningman.com.au