IET logo
 
IET
Decrease font size
Increase font size
Topic Title: UPS earthing and hazardous areas
Topic Summary:
Created On: 27 September 2017 10:27 PM
Status: Read Only
Linear : Threading : Single : Branch
Search Topic Search Topic
Topic Tools Topic Tools
View similar topics View similar topics
View topic in raw text format. Print this topic.
 27 September 2017 10:27 PM
User is offline View Users Profile Print this message



electronicsuk

Posts: 4
Joined: 25 September 2017

There are plenty of topics here about UPS earthing arrangements, but I don't think that one has been covered before. On our site we have a gas chromatograph that was recently damaged due to a power quality issue. I've suggested that we put it behind a UPS.

In theory it should be simple; it's a small load, about 100w, and it's the only load that would be fed from the UPS, so could conceivably be considered electrical separation (413.1.2). My understanding is that most modern UPS systems rely on the incoming supply to provide some reference between N and E, and if both poles of the incoming supply are disconnected, the load effectively has a floating earth.

I wouldn't normally be too concerned as small UPS systems are obviously designed by the manufacturer to operate with a floating earth on the output when running off battery. However, there's an extra layer of complication here as the gas chromatograph is an ATEX rated piece of equipment in a potentially explosive atmosphere. BS EN 60079-14 and -17 don't make any specific reference to UPS systems that I can see, nor does my CompEx guide book.

Maybe I'm just being over-cautious, but the idea of a floating earth in such an installation just doesn't seem right. Playing devil's advocate, if supply L & N were isolated at an upstream DB somewhere in the installation, the UPS would take over. As the N has been disconnected, there's no longer any current path between N & E as far as the UPS is concerned, as it relies on the two being connected at the substation. In an ideal world the load on the UPS would now be electrically separated, a fault between L & E should have no current path, no spark can occur and an explosive atmosphere can't be ignited. In the real world, certainly if the UPS maintains a connection between supply and load N when on battery, as some do, I'm not at all confident in such a scheme. Even capacitance between N and E of the supply cables (up to the point of isolation) could allow enough current flow to cause a spark in the event of a L-E fault at the load.

The best solution I have so far is to put a 1:1 isolation transformer between the UPS output and the load, linking N-E on the transformer's output, effectively creating a small TN-S installation. This 'new' earth would also be connected to the incoming supply earth to ensure no potential between the two. Even better would be to find a UPS with a built-in isolation transformer, but the only ones I can find are 120V American models.

Would appreciate any advice or pointers, especially from anyone with experience fitting UPS supplies to equipment in hazardous areas.
 28 September 2017 10:24 AM
User is offline View Users Profile Print this message



AJJewsbury

Posts: 16088
Joined: 13 August 2003

if both poles of the incoming supply are disconnected, the load effectively has a floating earth.

Do you mean a floating N rather than floating earth? (If PE is still connected, it's perhaps more like an IT earthing system). Any CE marked equipment designed for plugging into a normal socket won't be sensitive to "N" not being close to earth (since it will need to work safely in parts of the continent with unpolarised sockets, possibly with the two live pins connected between phases).

My understanding is that most modern UPS systems rely on the incoming supply to provide some reference between N and E

I'm not sure that's entirely right. As I understood it (I could be wrong) US-origin UPSs traditionally kept load side N connected to supply side N so as to keep a N reference during a power outage, however EN standards took the reverse policy and demanded that the load N be disconnected from the supply N when running on battery. The difference probably originates from the fact that US plugs & sockets are reliable polarised, whereas continental ones aren't. Plus of course leaving in-N and out-N linked is completely defeated if the user unplugs the UPS from the wall or the supply cable is severed/disconnected - and with the additional risk of filters within the loads tending to create a 115-0-115 voltage between N-PE-L you could end up with somewhat nasty voltages on the exposed pins of the supply plug (between N and PE). I thought APC et al were gradually moving over to the EN/international approach (at least for small plug-in UPSs). Other here have better knowledge of the specific standards, so may be able to clarify.

I gather that some every large UPSs still rely on the supply side's N-PE link - but the DPC for the 18th Ed of BS 7671 suggests that that's likely to be prohibited in future too.

The ATEX side is way over my head though - again perhaps others here will be able to help you with that.

- Andy.
 28 September 2017 12:54 PM
User is offline View Users Profile Print this message



electronicsuk

Posts: 4
Joined: 25 September 2017

Originally posted by: AJJewsbury
Do you mean a floating N rather than floating earth? (If PE is still connected, it's perhaps more like an IT earthing system). Any CE marked equipment designed for plugging into a normal socket won't be sensitive to "N" not being close to earth (since it will need to work safely in parts of the continent with unpolarised sockets, possibly with the two live pins connected between phases).


Yup, I could have phrased it better. Neutral floating with respect to earth. Good point about the design aspects of the equipment itself, though.

Originally posted by: AJJewsburyAs I understood it (I could be wrong) US-origin UPSs traditionally kept load side N connected to supply side N so as to keep a N reference during a power outage, however EN standards took the reverse policy and demanded that the load N be disconnected from the supply N when running on battery. The difference probably originates from the fact that US plugs & sockets are reliable polarised, whereas continental ones aren't.


That would make sense. Unfortunately, there seems to be a lack of technical information on the inner workings of most smaller UPS systems, even from the big players like APC, Eaton, etc. In that respect the idea of using a transformer on the output is quite attractive, as it doesn't matter how the UPS works internally, as long as it will handle the inrush current of the transformer at switch-on without going into overload.
 28 September 2017 07:22 PM
User is offline View Users Profile Print this message



AJJewsbury

Posts: 16088
Joined: 13 August 2003

Unfortunately, there seems to be a lack of technical information on the inner workings of most smaller UPS systems, even from the big players like APC, Eaton, etc.

I've got to sympathise there.

APC do have some interesting stuff hidden away on their web site though - e.g. http://www.apc.com/us/en/faqs/FA156549/

And if you're really short of bed-time reading they have quite a collection of "white papers" - http://www.apc.com/us/en/prod_...20Paper&Query_Type=10

- Andy.
 29 September 2017 02:36 PM
User is offline View Users Profile Print this message



gkenyon

Posts: 4981
Joined: 06 May 2002

It's not always the case that equipment powered from UPS that has been disconnected from its source supply earth / neutral is "floating" - because connections of other wiring (controls / data etc.) might provide a path to earth ... or alternatively, some of these wiring types (like Ethernet cables) might build up static charge if there's no controlled discharge path to earth ... not good in explosive atmospheres situation.

-------------------------
EUR ING Graham Kenyon CEng MIET TechIOSH
G Kenyon Technology Ltd

Web-Site: www.gkenyontech.com
 01 October 2017 03:59 PM
User is offline View Users Profile Print this message



kellyselectric

Posts: 191
Joined: 22 July 2016

Ok so I may be missing the point a bit here but why don't you just go for a really good filtered socket for your gas thingy if its been stuffed up by power issues then surely a good filtered surge surpressed socket will really help and you wont have to worry about floating neutrals etc
 01 October 2017 05:24 PM
User is offline View Users Profile Print this message



electronicsuk

Posts: 4
Joined: 25 September 2017

The APC white paper definitely makes for some interesting reading. I've also spoken with Eaton over the phone since my original post, and they say that their current philosophy is to completely disconnect both live supply conductors from the output/inverter on mains fail, so the UPS output is completely floating with respect to the incoming supply. Apparently they used to supply output transformers for exactly this kind of application, but they stopped producing them due to low demand. It's not like a small 1:1 isolation transformer is hard to come by, and I'm still leaning that way, as I'm more comfortable with the idea of earth being at a known potential.

Originally posted by: kellyselectric
Ok so I may be missing the point a bit here but why don't you just go for a really good filtered socket for your gas thingy if its been stuffed up by power issues then surely a good filtered surge surpressed socket will really help and you wont have to worry about floating neutrals etc


Unfortunately even a short interruption in power to the gas chromatograph can shut it down, and we don't always have staff on-hand who are familiar enough with the equipment to get it online again. If it's off for too long then it can take a while to warm up again before taking any readings, which causes knock-on effects for our other processes that rely on the chromatograph. It makes sense for it to be on a UPS if at all possible.
Statistics

New here?


See Also:



FuseTalk Standard Edition v3.2 - © 1999-2017 FuseTalk Inc. All rights reserved.

..