Data Center in close proximity to a 115KV Substation 1

[hugop973]

Is there a minimum separation recommended between a Data Center and a 115KV Substation. The Ground Potential Rise in an existing Data Center adjacent to a 115KV substation (wall to wall) is 809V, there's a concern about the EMI affecting the performance of the servers, routers, etc.
GPR of 809V is considered acceptable or high?
Any technical paper, report or investigation that can be used as a reference?

Thanks for your help

 

[umrpwr]

The short answer to your initial question is as far I aware there are no standards or recommended guidelines for the minimum distance for separation between a data center and substation.
It seems that you have raised a couple of issues in your post though.

The first question being whether the proximity of the data center to the 115kV substation will cause EMI/EMF problems under normal operating conditions. This a question of whether you may need to provide EMI/EMF shielding. A survey can be performed to measure the gauss levels and then determine whether EMF shielding would be necessary based on on the results which can then also be used in determining the design requirements for the shielding. There are consultants who specialize in this type of work and can conduct surveys, provide recommendations and design shielding if required. Other projects which I have been involved in though did not require shielding.

The second part of your question is regarding the GPR and risk to equipment within the facility. Having conducted investigations and performed analysis for similar projects with HV substatons >115kV and above in close proximity to the data center I would suggest you conduct a detailed grounding analysis of the entire facility which you may have already done or are in the process of doing. What may be an acceptable GPR to meet IEEE Std. 80 may not be acceptable for the operation of your data center. The GPR and transfer voltages are what you need to look at closely. Performing a grounding assessment and analysis will allow you to determine if the voltage gradiant will be an issue. My experience has been that you may need to make some changes and take additional steps to minimize any potential risks to the data center.

Unfortunately I am not aware of any specific papers or reports on this subject related to data center which are available to the public and I would not be not able to share specific information from some of my previous investigations. I would suggest though in addition to IEEE Std. 80 you take a look at IEEE Std 142 (Green Book), IEEE Std. 1100 (Emerald Book) which you probably have and also IEEE Std. 367 as a guideline and starting point.

I would also suggest that you may want to consider contacting a consultant such as Hood Patterson Dewar Engineers in Atlanta, GA.

http://www.hoodpd.com/commissioning-consulting.html

They are one of the most experienced data center commissioning firms and engineering consultants in the industry and a firm which I have worked with on numerous projects in the past including a couple with similar grounding issues as the one you mentioned in your post. They can act a consultant to you and have the knowledge and experience to be able to assist you in conducting a survey, performing analysis and providing recommendations. They specialize in grounding issues and data centers. In addition they also work with Dr. Meliopoulos at GA Tech as the distributor for his Smart Ground Meter.

http://www.hoodpd.com/grounding-services.html

http://www.hoodpd.com/smart-ground.html

 

[cranky108]

Why would a data center be any different than any other customer?

Besides there are ways around such problems.

 

[hugop973]

The client is also saying if there's more than 1 Km separation distance (from data center to substation) and GPR value of less than 300V with normal grounding constitutes a safe/normal environment for current standard off-the -shelf IT and communication equipment?
Who ultimately makes that decision?

Thanks for your help.

 

[cranky108]

Basically the step and touch potential in the substation, or to the substation fence is required to be in the safe area for employees, or the public out side the fence. And the ground grid should extend 3 feet out side the fence.

So how do you know the GPR levels? Are you guessing, calculating, or did someone tell you?

 

[hugop973]

They measured the GPR in the data center of 809V

 

[cranky108]

So maybe the data center needs better grounding.

GPR has to do with the highest phase to ground fault current into the impedance to a remote ground. The highest fault current will result usually from your lower voltages (darn current step up transformers), so you should be looking at the largest current voltege source, which most likely won't be 115 kV.

Many utilities use transformer impedance to limit fault current to some reasonable value. Like we limit our lower voltages to less than 10 kA, but if the data center uses transformers to step down the voltage more, the higest fault current voltage maybe one of those lower voltages. Do they have 4.16 KV? 480 V? something in-between?

 

[cuky2000]

Search for "Zone of Influence (ZOI)" and GPR for telecom industry. This is indirectly related with the distance between the sub and the data center.

With fiber optic connection provide good isolation of any unusual high GPR issue. Usually the phone company install isolation device based on the maximum expected GPR calculated in the substation.

 

[rcw retired EE]

Hugop973 - You mentioned "wall-to-wall" in the original post. Is that referring to the distance between the data center and the substation or is the data center wall immediately adjacent to the 115 kV substation wall?

If there is minimal separation, the two ground girds need to be tied together and precautions taken for transferred potentials for all conductive lines coming into the data center/substation.
Attempting to seperate ground grids a couple meters apart is not effective.

IEE Std 361-1996 "Recommended Practice for Determining the Electric Power Station Ground Potential Rise and Induced Voltage from a Power Fault" describes the Zone of Influence and how to calculate it.