I am doing an EMF / EMI inspection of an office building. I will only be spending
a couple hours per week inspecting the building until I can complete it. I will
send descriptions and pictures of what I find each week.
The main power quality complaints for the building are minor but include:
The 225 kVA transformer, (which
feeds 120/208-volt power to the building), is
very hot to the touch.
One tenant has several strip
surge suppressors that have the "fault" light lit.
The purpose of the inspection is to use the EMF given off by incorrect or defective
wiring to locate the wiring problems that are causing the power quality complaints.
As the problems are identified, an electrician will make the needed corrections.
This is a five-story rental office building, which opened in 1990 and is approximately
110,000 square feet in size. The power is fed through a 1200 amp 480-volt main
and the mechanical system is fed with a separate 600-amp 480 volt main. The
120/208-volt transformer is 225 kVA and there is a 45 kVA transformer for the
emergency power system. There are two large tenant owned UPS units, one of which
is fed from a separate outdoor emergency generator.
Before the EMF work begins, we need to look at the basics to see if everything
is correct. The initial inspection was in the main electrical room. The main
service bonding and grounding was correct. Now we move to the 225 kVA transformer.
The transformer secondary has
two parallel feeder conduits and the temperature on the top of the transformer
is 138 degrees F.
Note: To check the outside surface temperature of the transformer, I use a handheld
infrared thermometer (cost about $180 from Graingerís). I aim the infrared into
the transformer up underneath the top. I have found that anything over 100 degrees
F is another symptom of a possible electrical problem.
The current on the grounding
electrode conductor is whopping 25.5 amps. The goal is to correct all the "net
currents" in the building and reduce the 25.5 amps to about 2 amps or less.
The metal raceway containing
the grounding electrode conductor is bonded at both ends to the conductor.
The secondary feeders (EMT and
Flex) do not have an equipment grounding conductor terminating to the X0 terminal
on the secondary. At this point, the only path for ground fault current back
to the power supply (transformer) is the feeder conduit, (which includes flex).
I do not know if the transformer is grounded
correctly in the transformer or if the transformer case is properly bonded.
I instructed the electrician to schedule a shut down of the transformer as soon
as possible to:
Properly bond the transformer
case to the neutral.
Install equipment grounding
conductors in the secondary feeders or to add a bonding jumper for the 6í flex.
Ground the transformer to building
Tighten all electrical connections.
Since this shutdown needs to be done off-hours
with advanced notification to the tenants, it may take a week or two to schedule
Note: A walkthrough with a triaxial Gaussmeter confirms what the transformer
already told me, there are "net currents" present in the building.
Next week I will start to identify the circuits causing the "net currents"
but the electrician will not make any corrections until the transformer work
Please note that EMF inspection should only be part of a complete electrical
maintenance program. The program should include regular thermal imaging inspection,
scheduled tightening of connections and regular inspections by qualified electricians
and electrical inspectors.
Comment No. 1
Mike, The "hot" transformers that
we have investigated are due to 3rd (or zero sequence) harmonic current produced
by electronic loads. The 3rd harmonic current add together in the neutral (possibly
overloading the conductor) and ends up circulating in the transformer primary
delta winding. Based on the total harmonic current a derating factor should
be applied to the transformer. It seems as though this survey should be looking
at the harmonic current as well as the grounding issues.
Comment No. 2
Is "net current" a term some use instead 3rd harmonic current?
Chris Fate, Snohomish County PUD
The hot transformer may be overloaded or more
than likely suffer from harmonics. With UPS loads that do not have an input
filter (most UPS systems not all) the harmonic content can be significant. Does
Ed have a harmonic test instrument? Check the neutral current for harmonic content.
What loads are running on the transformer?
When transformers are shipped, the X/O bond is not made and I have seen a number
that were never bonded by the electrician. I will not get into the problems
that can cause.
Has Ed measured the resistance of the building grounding?
I quit blaming everything on harmonics a long
time ago. Bonding and grounding errors and neutral to ground shorts are where
I concentrate now.
Just like a 100-degree day is the best way to find the weaknesses in your refrigeration
system, harmonics are great at pointing out the weaknesses in your electrical
I use the term "net current" two different ways. One-way is: the lack
of, or excess of, current on a complete set of conductors Take a branch circuit
coming off a breaker box for instance. If you clamp all the conductors together
you should get a zero reading. Today as an example, I clamped a circuit where
there is a neutral to ground short in a receptacle and the reading was 3.5 amps.
So the 3.5 amps was the net current and the 3.5 amps was neutral current running
on the steel of the building and not returning through the grounded conductor.
The other was referring to a single conductor that has current on it and normally
shouldn't. For instance, the water pipe ground at your home or the transformer
to building steel grounding conductor.
Note: I work with a local electrical testing
company that does the thermal imaging, ground resistance testing and power quality
testing. Those tests are a standard part of any electrical survey. I am detailing
only the EMF portion, as not many are familiar with it.
Edís Report No. 2
I am a little short on time this week but
I had a second look at the property to continue my inspection. I have found
fourteen (14) 120 volt branch circuits on the second, third and fifth floors
that have "net currents" greater than .5 amps. I did not do the first
and fourth floors yet.
As the building has a small 24/7 operation
on one of the floors, we have not been able to pin down the tenant for a time
for the transformer shut down. Next week I will be working with an electrician
to clear the "net current" circuits and inspect the other floors.
Note: There were 4.25 amps of "net current"
on one circuit. Using the formula B=2I/R, (where "B"= flux density
read in milligauss, (mG), "I" is the "net current" and "R"
is distance in meters), means that along the entire length of the cable run
in the ceiling, the EMF reading given off by this cable alone, in all directions,
is 8.50 mG one meter away from the cable. 4.25 mG two meters away.
I hope this shows the significance of using
the Gaussmeter for a quick walkthrough. Also, if you know the distance from
the source of the "net current" you can take the Gaussmeter reading,
invert the formula, and estimate the "net current".
Edís Report No. 3
Sorry I have had nothing to report on the
EMF inspection. I did not work on it the past few weeks waiting for the transformer
shut down that was scheduled for last Thursday night. The 24-hour tenant canceled
the shut down at the last minute.
The plan was for the office to shut down.
While reviewing the equipment in the computer room it was discovered that other
offices use the same computer server, and it cannot be shut down. They have
a large UPS unit but not a generator. We also discovered the UPS unit feeds
two 10-ton A/C units.
So before we can fix the transformer, we need
to fix the wiring of the computer room power. In the mean time we will precede
continue with clearing the net currents.
I see this all the time in critical computer rooms. The computer people know
their equipment but not everybody understands how their electrical supply does
or should work.
Edís Report No. 4
Things have slowed down a little so
I was able to get back to the EMF inspection I started last month. As I mentioned
earlier (Report No. 2) I had high Gaussmeter readings on the fifth floor and
we identified the circuits that were causing the EMF's. Yesterday we started
to trace them and all I can say is what a mess!
This floor was remolded two months ago and about 8,000 square feet of cubicles
were installed in a large open area. Each section of six cubicles has 3 circuits
fed through a single whip terminated to a wall junction box. The whip has three
12 gage hot wires and one 10-gage neutral plus a 12 gage grounding wire. Each
cubicle then has three receptacles, one marked for each circuit. Since there
is only one neutral, the intent was to wire the whip to a three circuit one
neutral house feed, each circuit on a different phase.
While I was there I learned that this tenant has been experiencing frequent
monitor failures since moving in. I'm not surprised.
The electrical contractor used a two circuit
home run feed with one same size neutral. Then at ceiling junction boxes dropped
one individual circuit to feed two of the three cubicle circuits and brought
another circuit from a different home run, (sometimes from a different panel),
to feed the third circuit in the cubicles. There was no regard for what phase
anything was on and all the different circuit neutrals are tied together in
the junction boxes.
This tenant is in operation from 7:00 am to 11:00 PM 7 days a week, so the contractor
will have to work the midnight shift to straighten this problem out.
Ed Chavern, Chavernet@aol.com