Today is...
Saturday, May 26, 2018
Welcome to, the global online
community of automation professionals.
Featured Video...
Featured Video
EtherCAT with CTC’s master lets your multivendor network play well together...
Our Advertisers
Help keep our servers running...
Patronize our advertisers!
Visit our Post Archive
Separate Earth for Intrinsically Safe Instruments
Is a Separate Earth for Intrinsically Safe Instruments required?
By Hamid Sarukhani on 16 August, 2008 - 12:23 am

Dear All,

As far as I know we can not connect the earthing connection of Intrinsically Safe instruments to common clean earth where the shield of other instrument cables and other electronics are connected. In other words, Intrinsically Safe Earth should be used for IS signals only and not to be connected galvanically to other earth loops.

Is the above statement correct or we can share the clean earth with Intrinsically Safe signals?


There is no such thing as "intrinsically safe earth." All grounding points must be connected to a common earth grounding point. That said, you MUST be sure that your grounding point is actually at earth ground potential. Building steel is famous for not being a suitable grounding point.

Even copper bus bars that make up a building buried earth grounding grid may not be at true earth ground potential. The only way to validate your earth grounding point is to test it with a megger. If you have a high ground water table, and your grid is submerged in the ground water, you are OK. This is rare in most non-coastal locations.

Dick Caro
Richard H. Caro, Certified Automation Professional, CEO, CMC Associates,
2 Beth Circle, Acton, MA 01720
Tel: +1.978.635.9449 Mobile: +.978.764.4728
Fax: +1.978.246.1270
Dick Caro

Dear Dick,

I was working in a gas plant Construction project with ENI (Agip) where the DCS was by Yokogawa.

In the earthing principle by Yokogawa it was clearly shown that there were separate earths:
1.IE Electric Ground (Instrument Earth)
2.IPE Primary Earth Ground (Instrument Protective Earth)
3.PE Plant Earth Loop
4.ISE IS Electronic Ground (Intrinsically Safe Earth)

The first 2 earths had their own loop andthey were galvanically connected to PE while ISE was totally separate and was not connected to PE.

Asking me, there must have been some sort of logic behind it, assuming Yokogawa and ENI do not do useless practices.


1 out of 1 members thought this post was helpful...

Yokogawa was being very conservative. In the ideal, a grounding grid submersed in ground water provides a zero resistance between all grounding points, and cannot form ground loops between grounding points. If the grounding grid is not in ground water, then different grounding points may actually have resistance between them. If there is any resistance between grounding points, then ground loop currents may happen.

The Yokogawa practice is designed to keep ground loops formed by any imbalance in the electrical power wiring from finding a ground in the instrument system, if the ground points have any resistance between them. This is the situation for dry earth grounds, but is not necessary if the grounding points are all submerged in ground water.

Dick Caro

First, check the requirements for earthing in an intrinsic safety system in your locally-applicable standards. The IEC 60079 series requires a resistance of less than 1 ohm between the IS earth of a barrier system and the neutral point of the relevant AC power supply system.

This is to ensure that in the event of a short between the AC supply system and the low-voltage feed to the barriers (on the safe side) the protection will operate and the voltage rise of the IS earth system during the fault will be within bounds. This requirement is usually met by a pair of large-diameter earth connections between the IS earth and the neutral point or the earthing point for the power supply neutral.

The earth loop resistance can be checked by disconnecting one of the cables at one point and measuring the resistance across the break. IEC60079 also requires that IS cable screens be connected to the barrier earth. Secondly, the IS earth is usually also connected into the instrumentation system and needs to be connected to the instrument earth or common line. For EMC reasons, there must be only a single connection between the instrumentation earth system and any other services.

The best way to do this is to construct a tree with the root being a single connection to earth common to all - power, IS, instrumentation. Power connections are one branch off this: instrumentation and IS another. The instrumentation and IS systems are separated at a single point.

Instrumentation earths should be further branched into digital and analogue sub-systems. Any connection to the rest of the world by a physical connection into a water table or well or a rod driven into the soil does not really change this - unless you are dealing with a power station, where earth faults on remote lines need to be taken into account. The metalwork on a site - including enclosure boxes, motor frames, and pipework - needs to be bonded together to form an equipotential mesh and also connected to the site earth point.

However, there may be a large voltage difference on points connected to this equipotential mesh, and large currents circulating around it. Earthing is a topic fraught with mystical beliefs such as keeping instrumentation and power earths totally segregated yet physically connected. Go back to first principles and mak sure the requiremenst are defined and satisfied.



what if we have the same earth (Is, Non IS and PE) and that earth is not proper? Please appreciate that the lightning arrestor is also connected to same earth. Any discharge from the lightning if not finding a conducive earth path will affect the instrumentation system. In such a situation, don't you feel that separate earth will protect instrumentation because the Plant earth is designed to carry such discharge because the PE is where the discharge goes? Provided while in construction we maintain sufficient resistance between the plant and instrument earth once disconnected. We had one control expert from Gas turbine company who said that the resistance should be more than 5 K Ohm (Possibility of loop current is minimized).
May be i am wrong. Compressor station where it was done sincerely, did not have any ghost trips and others had their share of ghost trips.


By Bruce Durdle on 11 July, 2012 - 4:58 pm

Hi Anil,

Lightning protection has its own issues. Any lightning surges have the potential to get into all sorts of funny places and don't necessarily follow the usual rules. For example, I have been told of a case where a large vessel containing an unstable liquid and its vapour was ignited by a lightning pulse passing through the vapour although the vessel was made of fully-welded metal. The issue was traced to the reactance of the path between the top of the vessel and the side wall, which involved a right-angle. Rather than follow the metal path, the lightning surge partly flowed through the vapour, causing it to decompose.

The IEC IS standards require that cables and wiring are capable of passing an insulation resistance test at 500V - this would usually require that there must be at least 1 Mohm resistance between any two conductors in an IS system (this includes between the 2 conductors in a pair, and between a conductor and its associated screen). So the insulation resistance between an IS conductor and any other conducting element must be at least this.

The other requirement is that the resistance between the IS barrier earth connection and the neutral point of the associated AC power supply is less than 1 ohm. This is NOT a generic non-specific earth point associated with the power system - the reason for this requirement is to ensure that any power supply faults feeding in to the "safe" or non-IS side of the barrier cause the internal fuses to blow and are not propagated through to the circuit in the hazardous area.

While the lightning protection earth and any external power feeds (via MV or HV transmission lines) must be connected to the "general mass of earth", there is no need for the internal systems to be so connected. Unless you have a very wet sub-soil (which will pose other problems) it is in fact very difficult to meet the requirements for a suitable connection and an earth connection must be specifically designed for each site. A solid copper connection at a single point between sub-systems is a much more controllable and reliable earthing method.


Hi Bruce,

Is there any separate electrical symbol for Intrinsically Safety Earth? Because PE and IE has separate symbol.


By Bruce Durdle on 17 November, 2014 - 6:49 pm

Not to my knowledge - but there is no reason why you can't adopt one for your own site standards if you feel it's a good idea.

By Susi Sukaisih on 27 August, 2009 - 8:54 pm

Dear Dick,

I'm new in Instrumentation, I need some reference about ISE Bar, how to Install it, As per Mr. Hamid review, its make confuse, coz ISE Bar, is it should be stand alone or should integrated between building and field?

Please give me some information.

By Mutlu Gundiler on 4 April, 2010 - 9:52 am

Dear Hamid,
Could you please provide me (or provide a link) the Yokogawa Spec which you have referred as below

"In the earthing principle by Yokogawa it was clearly shown that there were separate earths:
1.IE Electric Ground (Instrument Earth)
2.IPE Primary Earth Ground (Instrument Protective Earth)
3.PE Plant Earth Loop
4.ISE IS Electronic Ground(Intrinsically Safe Earth) "

Kind regards
mgundiler [at]

By Bob Peterson on 17 August, 2008 - 1:07 pm

all the separate grounds are typically connected together at the service point.

System suppliers often insist on separate earths thinking it will eliminate interference but at the end of the day all the earths should be tied together.

Bruce's excellent reply covers most aspects of the subject but I can send a copy of an old paper written by Chris Towle "A definitive guide to earthing and bonding in hazardous areas" to anyone that wants more reading material and contacts me off-line.

syates at mtl-inst. com

Thanks Steve,

I just downloaded the file from MTL.

In fact there are a dozen of nice docs there where you can download by free registration at MTL.

Perhaps some one can clarify for me then.

As I recall, IS with a zenner barrier requires a star earth but if using galvanic isolation it is not required. I seem to recall this being a problem for offshore platforms where a good star earth was a pain to provide for zener barriers but they were much happier with galvanic isolation.

As I recall, it only needed an additional certificate for the galvanic isolation and no design changes to the sensor.

But I speak without knowledge here so would be glad to be corrected.

In my experiences in the Petrochemical Industry

Depending on the situation it is preferable to use a High Frequency Choke to separate the IS earth and the safety earth. This prevents interference like noise being induced into the IS system. a spark gap can also be used as additional protection from high voltages being induced into the system.


Can you able to provide spark gap selection between instrumentation earth to electrical earth.. If possible.