STEP AND TOUCH POTENTIAL TESTS ON LARGE SYSTEMS

[xxjohnh]

It seems this is becoming more popular. Has anyone had good results on large generation plant ground grids? 1000 MVA and up. I see more requests for this now, but have never done one. How did it compare to the calculated values. I suspect most grids are overdesigned by 50% as people use IEEE 50 parameters and treat the whole plant as a substation.VDE also has some requirements.

 

[rcw retired EE]

This test was done on a partially completed plant to prove to the Safety Authority that it was safe to energize the 230 kV switchyard even though parts of the power plant were not complete. (Their concern was the design included the beneficial effect of 6" of gravel on the allowable step and touch, but no gravel was present in a large part of the plant. Also, the inspector asked for verification of the calculated current split between the grid and overhead ground wires.)

A consulting company specializing in these tests brought in a 55 Hz generator and recording equipment that could measure just the 55 Hz voltages and currents. The incoming transmission lines were grounded at the far end and the generator was connected to simulate a fault in the 230 kV plant switchyard.

Measurements were made on the overhead ground wires and step and touch values were measured at selected points around the site such as entrances to fenced areas and steel buildings where exposure is high and at areas where the grounding software indicated higher touch voltages.

Test results were analyzed by the consultant and measured values of grid impedance, current splits and projected step and touch voltage were given. The design was proven to be conservative.

It was an expensive test, but the data was worth it. I would like to go back and redo the test on a completed facility to compare to the design, but no one wants to pay for it.

I don't think it would be cost effective to measure all possible step and touch potentials in a completed plant. But doing a selected few to prove the design may be worth while, especially if it is difficult to perform a fall of potential test to measure the gird resistance.

 

[xxjohnh]

This would be worthy of an IEEE paper. How far was the generator from the Substation? Many folks confuse IEEE80 and IEEE 665 and you see grids in power plants the same size as the substation. What standards were your grids designed to, sounds like IEC?

 

[rcw retired EE]

xxjohnh - This Canadian power plant grid had to be designed to IEEE 80 and meet the Ontario Electrical Systems Code that refers to IEEE80 plus has specific step and touch potential. Our designs usually follow IEEE 665 backed up by computer simulations using SES software. The software calculations are based on IEEE 80.

The Safety Authority inspector was a very knowledgeable engineer (PEng with an MSEE) who was writing his PHD thesis on how IEEE 80 based computer programs incorrectly calculate step and touch potentials, especially at the interfaces between two surface treatments like asphalt road and switchyard gravel, or gravel and concrete foundation. That was the major reason for a special test to validate our design.

I did not get enough data for an IEEE paper, but I’m working on it.

 

[RNRPE]

I agree, Step and Touch Potential Measurements are becoming more common. It's really the only way of determining if a ground grid was installed according to design or to monitor the integrity of the grid over time.

I have found that many firms and utilities cannot conduct step and touch potential measurements properly, since establishing the test criteria often requires knowledge of lead coupling and interference, as well as having a good understanding of grounding system design.

IEEE 81.2 and HD 637S1 are two standards I use when performing ground impedance and step and touch measurements, but they alone do not answer all the questions. Software modeling mybe be required to fully understand measurement result.

 

[pwrtran]

Does anyone know any company in Ontario can provide step/touch potential field survey by a high current injection > 100A? The background is:
I worked on a substation grid design in the past. The entire site was a swamp and was filled by dumping 8'-10' huge rocks. The sub was of course build on the rocks and there was very minimal soil layer on top of the rocks. One can image how poor the soil resistivity would be. There was no easy way to dissipate fault current in the main grid area other than bring the current to a remote grid. The distance between the main and the remote is about 150m and the grids was interconnected by 2x556 acsr on the O/H run. The ESA plan adviser didn't recognize using the combination of the main and remote grid to reduce the GPP. He insisted to control the step and touch based on the main grid ONLY because his argument was the double 556 interconnected ACSR could be broken by a pole hit and the remote grid was lost. That would be a triple contingency - 1) you must have a car hit the pole; 2) the pole must be down and the acsr is torn broken; 3) then followed by the primary side L-G fault in the sub plus someone was touching a metal parts in the sub at the moment. He didn't trust any the calculated current splitting unless you do a current injection test. He insisted to do a high current injection step/touch field survey if we use the combination of the main and the remote. Eventually, we ended up with burying tons of copper in the main and paved the whole site asphalt.

 

[cuky2000]

I worked in a project that achieved good approximation between the calculated and tested step and touch potentials values.

This project consisted in a combined cycle power plant located in a high density populated metropolitan area that supply power to a local indoor switchyard interconnected to a nearby outdoor area substation via underground MV superconductor cable system and 115 kV solid dielectric feeders. The outdoor substation was interconnected with 230 kV and 345 kV remote stations via pipe type feeders. A step and touch potential test was measured using the Smart Ground Multimeter (SGM)

http://www.hoodpd.com/products.html

The following results were compared favorable as follow:
• Outdoor substations: the test and calculated values were consistently close in three randomly selected locations. The grid model was also accurately modeled.
• Indoor switchyard: the calculated results were conservative in the safe side. The concrete slab behaves as a quasi equipotential surface. Grid model and earth/concrete resistivity was difficult to model accurately because of many UG metallic bodies (rebar, water main, gas pipes,conduit,tunnels,etc) and moisture variation of concrete slab and ground.

I am not sure regarding the comment

IEEE Std 80 incorrectly calculate step and touch potentials

For sure, the IEEE std 80 does not explicitly address the calculation on interfaces between two different surfaces and is silent regarding recommendation of footwear resistance. In fact, the permissible step and touch potentials is conservativelly calculated considering a person with bare food standing in a wet surface with a body resistance of 1000 Ohms.

_{The good news is that there is a way to calculate the allowable step and touch potential for two different surfaces with a good degree of accuracy. However, this could be very controversial, since most engineers prefer to be in the conservative side. The issue at the end is safety vs. economics.}

 

[RNRPE]

Cucky2000 is correct, but their is software and test equipment to model the human and footware aspect. I will be demonstrating a test equipment on behalf of Omicron Electronics in June on a Hydro Quebec facility. The nice aspect of this equipment is you can use a transmission line as the current injection lead and usually achieve a high current. It's hard to achieve 100 amps unless you can impress a very high voltage between the two points. Omicron also has a Fast Fourier Transform meter that is frequency selctable to read the frequency of the injected current. The meter has dip switches which can also insert a 1000 ohm resistor to represent the average human body resistance and also insert another resistance to simulate, for example, EH-rated footware.

It's expensive equipment, but it works very well for this application.

With respect to software, I like using CDEGS from Safe Engineering Services in Laval, Quebec, Canada. Their grounding software can simulate any resistance and is a good tool for understanding fault current distribution, which is an important factor in understand step and touch potential measurements.

 

[RNRPE]

pwrtran: I guess I did not address your underlying question. I don't know any company that can perform a high-current S&T potential measurement for this site in Ontario, or in North America. However, if you provide me details about the site and it's location I can specify the equipment and protocol. Will this help?

 

[rcw retired EE]

pwrtran - Check out Kinectrics Inc, 800 Kipling Ave, Toronto, Ontario.

They did the test I mentioned in previous posts. They have a good realtionship with the ESA people and can provide good advice on whther the design is safe or not.

 

[pwrtran]

Thanks RNRPE and rcwilson. We did the fall-of-potential test for the main and remote separately and they both agreed with the calculated value, which means the main grid is capable to control the Step/Touch under the maximum fault current without the presence of the remote and the remote takes care of the GPR. ESA was happy with it.