first time ive seen this

Status
Not open for further replies.
please do not dismiss me because of my age i did not get to where i am today just because of my cute looks or family prestige, like my sister who also graduated higher education i have also recently done so. i have two masters degrees one in electrical engineering and one in information technology data center management. i have also recently moved to the united states from Sapporo japan. now i will get to the point.
the public utility here is Puget sound energy.
i am used to dealing with utility company's on power quality issues and other things. but this has me flustered.
the site i am working is a small data center, aprox 1,870 Sqf served by a 200A 1 phase 120/240 volt service via overhead drop.
the site has all new wiring and load-centers (cuttler hammer)

the poco says that the service drop is sized properly but, why is a load as small as 7.8 amps able to bring down one leg of the service the service drop is 2/0 alum. the feeder between load centers is 3/0 copper with a length of 54 feet +- 3 feet.
with a load of 11.7 amps applied to one leg i can bring the voltage as low as 109 volts on that leg. this is mirror on the opposite leg if i apply the load to it . moving on to the 100A sub panel feed from this 200A mains i can cause the legs to swing a symptom of a bad neutral that is also present at all the load centers. this service has been replaced from the service drop to the meter base clean back to the sub panel. 1 year ago and i have myself inspected the entire system and found all lugs and components to be in proper order with no damaged stranding or other faults that would cause the, unbalance to be present or such a high fop. am i correct in my conclusion that the overhead line between the pot and service head is bad or that the transformer itself has a bad connection inside it? the poco has remade the splices on this service drop 3 times with the same issue remaining.
 

GoldDigger

Moderator
Staff member
Location
Placerville, CA, USA
Occupation
Retired PV System Designer
If putting a balanced load of 10A on both phases at the same time does not cause a shift or high voltage drop on either phase, you have confirmed a neutral problem.
If you measure the voltage shift at the main terminals you can be sure that the problem is upstream of that point.
POCO re-doing the splices without going all the way back to the center tap of their transformer will not solve the problem.
POCO needs to measure voltages on their side while there is an unbalanced load on the service. If they do not come out with a load box, they are not doing their job.

Sent from my XT1080 using Tapatalk
 
Location
MA
It's pretty common for POCO's to come out and change all connections on fluctuating voltage/flicker calls without really troubleshooting. It does fix things nine out of 10 times. Most of the time there is nobody at the address and/or they have no way of adding load so all the connections get replaced as a guess. You should call again and politely ask them to try again after you check at your main like GoldDigger said. I doubt your service is 2/0.. more likely #2. Almost all POCO feeds are undersized technically. You will find that a lot in the US.
 
the client basically dropped this all on me to fix and, yes ill climb the dam pole myself if i must:rant:. the poco has placed a recording volt meter on this service at the meter base 4 times. then claimed the fault is on the clients side of the meter. as an ee i can smell the load of bs on that a mile away. first because the rvm was put on that service twice prior to the electrical contractors at south sound electric coming out and replacing everything from the top of the service drop pole to the distribution panels inside. this was done at a cost of $7,800 usd and i have myself inspected their work it is perfect not in just the quality, but they did it above the min req by the nec. the pq and noise issues are still present after the job was completed. i have used the rvm the poco uses they are not capable of determining what side of the meter base the issue is. because they are using the service neutral as their zero point not a separate ground reference point. if anything the data it gathered in this configuration would tend to point at them. they just do not want to come out and replace the overhead service drop from their pole. i am not an electrician but it is my job to protect the data center and its contents by any means at my disposal. from damage of an electrical nature. that dam wire coming in here is to small and ill drag my ass to platt electrical supply and buy a spool of 4/0 copper and 3/0 support neutral messenger. and replace the dam pos myself and bill them for the cost of the job.
 

Smart $

Esteemed Member
Location
Ohio
Indeed.. and wire manufacturers. Just pointing out that coming from somewhere else where OP might not see a lot of overhead wire, he may wonder why the utility service is so small.
The target of wire manufacturer's published information is the consumer. Much is through compliance with NRTL standards. Ultimately, if the utility service is too small, voltage drop to below their standard will become apparent. In the case of the OP, this would typically show for line-to-line loads as well as line-to-neutral loads... i.e. with same-size, non-compromised conductors and "solid" connections.
 

jim dungar

Moderator
Staff member
Location
Wisconsin
Occupation
PE (Retired) - Power Systems
.... 200A 1 phase 120/240 volt service via overhead drop.

Has anyone checked the transformer?

I once worked with a machine shop that was in a barn type structure in 'farm country'.
When our loading hit about 150A, the utility voltage couldn't keep up. Finally got someone to climb the pole, where they noticed the installed transformer was only 15kVA. The utility reworked their math, put in a larger unit (size unknown) and our PQ problems became manageable.
 

petersonra

Senior Member
Location
Northern illinois
Occupation
engineer
the client basically dropped this all on me to fix and, yes ill climb the dam pole myself if i must:rant:. the poco has placed a recording volt meter on this service at the meter base 4 times. then claimed the fault is on the clients side of the meter. as an ee i can smell the load of bs on that a mile away. first because the rvm was put on that service twice prior to the electrical contractors at south sound electric coming out and replacing everything from the top of the service drop pole to the distribution panels inside. this was done at a cost of $7,800 usd and i have myself inspected their work it is perfect not in just the quality, but they did it above the min req by the nec. the pq and noise issues are still present after the job was completed. i have used the rvm the poco uses they are not capable of determining what side of the meter base the issue is. because they are using the service neutral as their zero point not a separate ground reference point. if anything the data it gathered in this configuration would tend to point at them. they just do not want to come out and replace the overhead service drop from their pole. i am not an electrician but it is my job to protect the data center and its contents by any means at my disposal. from damage of an electrical nature. that dam wire coming in here is to small and ill drag my ass to platt electrical supply and buy a spool of 4/0 copper and 3/0 support neutral messenger. and replace the dam pos myself and bill them for the cost of the job.

Are you properly trained for this kind of work?
 
the pot is a 25kva serving the data center and a home no other loads on it and i told thease jokers there are three failure modes on a pot not two
they think a pot works or it don't work well that's bs Tran failure modes are
1 open circuit and no boom
2 insulation failure and large chance for a classic Tran bomb
3 sec lugs get lose inside the tank and the pq goes to shit no signs of overheating cause of the dissipation from the connected service leads in free air
 

junkhound

Senior Member
Location
Renton, WA
Occupation
EE, power electronics specialty
Since I've dealt with PSE and have know a few PSE engineers, I'll throw in a few thoughts.

with a load of 11.7 amps applied to one leg i can bring the voltage as low as 109 volts on that leg


Am sure you have calculated that R=11/11.9 = about 1 ohm. And that the dissipation from wherever the drop occurs is close to 130W (IF resistive!!).

Don't know which county you are in on PSE, but PSE subcontracts almost all their feeder drops out to Potelco. My own house has a 50 foot overhead #2 Al service drop, common for here. Even with my heat pump running at 20A my outlet voltage is 120V. My house is server by a 37.5 kVA xfmr along with 5 other houses.

Potelco and PSE are well known for having the minimum sized xfmrs, a Habitat project with 5 units (electric kitchen but gas heat and water heater) was served by a single 25 kVA xfmr. When asked about that, PSE said the 25 kVA was rated for 150% overload, etc.....

Now, you said the load is a data center. Have you looked at the harmonics of the loads with a spectrum analyzer or a scope? What PF are the data center loads? If a lot of cheap full wave single phase rectifier from ends, lots of mischief can happen.

Back in the '90s, the major employer in the area started installing desktop computers for every employee. They all had single phase full wave rectifiers at the front end, and no PFC circuits or even PF correction caps. Part way thru the install, the main distribution transformer went up in smoke........

You said 11.7A gives you a 11V drop. Have you measured the before and after PF and phase???? Is the data center already drawing 150 A when you add the 11.7A?? 200A *240 V = 48kVA. If you ARE drawing over about 40 kVA, it is well possible that the reactance of the little 25 kVA transformer is giving you the drop? All the data is needed, not just implying that there is no load and putting 11.7A on line drops the voltage 11 Volts.

If there is a 1 ohm or so resistive line loss, then at 100 A you would be heating something up with about 10 kW!! If real, PSE would be on your case in a flash for stealing power (very closely monitored with the electronic metering in place for the last few years).

So, my guess is that your data center needs a distribution transformer with a lower leakage reactance.... get your real loads including power factor and harmonics and talk to PSE again. Having recently had PSE service installs, I know that about the ONLY thing they look for in sizing distribution transformers is flicker -- no big 200A motor start, or installing a 150A tankless water heater without paying a premium for a bigger transformer.

Post your real loads and maybe someone can lead you in the right direction without having to guess the data center loads.


BTW, you can get yourself and your employer into serious legal trouble in WA state climbing a power pole and installing your own service drop - not like in Japan where you can screw a circuit breaker to a wooden post in the kitchen without a box even and run an appliance.
 

ceb58

Senior Member
Location
Raeford, NC
the pot is a 25kva serving the data center and a home no other loads on it and i told thease jokers there are three failure modes on a pot not two
they think a pot works or it don't work well that's bs Tran failure modes are
1 open circuit and no boom
2 insulation failure and large chance for a classic Tran bomb
3 sec lugs get lose inside the tank and the pq goes to shit no signs of overheating cause of the dissipation from the connected service leads in free air

First off welcome to the forum, secondly out of respect for the owner and supporter of this forum we do our best to keep the profanity out of the post. That is but one thing that sets this forum apart from the others.
Now to your problem. It has been said you need to go back to the utility and request an engineer and a service tech. to come and place "the beast" on the meter base and load bank their lines. Chances are they will find the problem on or in the transformer.
 
the beast as it is named has been here sevral times already and ive already had several discussions with their pq engineer she knows her stuff. the datacenter has a pf of .78 which is not bad for a 4kw 8u rack with redundant pem's. this client is not a tier 1 data center it is a privately owned cloud data center. the total load on the a&b phases are L1 peak 54.87A L2 peak 55.27A this data center is in a residence the loading figures are real power taken at the main bus in the service panel. the second load is the residence across the street. they are not home most of the time. but it is an all electric residence. but unless they mis labeled the pot... a 25kva pot should be able to supply two residences with 200A services with a fop of under 1.3% on a #2 al overhead spanning a max of aprox 85 ft if you believe the way pse calculated their line size. free air or not resistance is resistance if the wire is undersized it will have outrageous losses regardless of how well the heat is dissipated!
 

suemarkp

Senior Member
Location
Kent, WA
Occupation
Retired Engineer
Is this data center in a residential area, and does PSE know this is a data center? They do skimp on service drop wire, as they ran 1/0 AL for my 400A service. They are willing to go larger, but you have to may for it. I agreed to do that (asking for 4/0 AL), but it wasn't on the Potelco trunk when they got here and they never charged me for it... They did at least install the 50 KVA transformer I paid extra for. If they know its a data center and not a dwelling, perhaps they won't skimp quite so much.

If you're just drawing 10A of load and the service voltage is dropping to 109V, you have a major problem (would think its a bad transformer, or a bad connection at the transformer). You should not be seeing anywhere near 15V of drop on properly terminated #2 AL unless you're pulling 200A through it.
 
Location
MA
The target of wire manufacturer's published information is the consumer. Much is through compliance with NRTL standards. Ultimately, if the utility service is too small, voltage drop to below their standard will become apparent. In the case of the OP, this would typically show for line-to-line loads as well as line-to-neutral loads... i.e. with same-size, non-compromised conductors and "solid" connections.

I was never disagreeing with you. I agree with everything including the last sentence. Just saying that one may look and think the wire is too small and have documentation to back it up when in reality #2 for 200 amp is perfectly fine.

Sometimes all these jobs take is a simple brushing of the transformer neutral lug and the wire. That type of fault can be hard to find since the wire looks good on the outside and it's making slight contact so you will not get wide open readings with a Beast or similar load. Utilities are not in the business of guessing when it comes to changing transformers. If the data says that it's not on their side, they are going to want you to make an effort/replace some things on your side before they change the expensive stuff. Now that you have demonstrated how much work you've done on your side, they may be willing to replace the transformer and see what happens. The same scenario happens a lot where I am. Sometimes things have to be narrowed down to an educated guess. That doesn't make everyone besides yourself an idiot and a sloppy worker.

Sound like your frustrated, but acting like a jerk will not help your case. If you do end up trying to run 4/0 up a pole by yourself, please video tape it..
 
chasing a rat up the pole is my last resort. and yes i will...if i do post it on youtube. the only advantage of being related to royalty.. is i get the diplomatic status..fixing things yourself is also the cause of major fits in japan but yes when my last and only option is to fix something myself.. i do so and then i stick the resulting legal bill on the entity that acted in such a way that i was made to take that option. but like i said i dont climb poles unless i am out of options, then it gets expensive for the utility because my sister who is an attourney will find a way to stick them with the entire bill.
 

junkhound

Senior Member
Location
Renton, WA
Occupation
EE, power electronics specialty
chasing a rat up the pole is my last resort. and yes i will...if i do post it on youtube. the only advantage of being related to royalty.. is i get the diplomatic status..fixing things yourself is also the cause of major fits in japan but yes when my last and only option is to fix something myself.. i do so and then i stick the resulting legal bill on the entity that acted in such a way that i was made to take that option. but like i said i dont climb poles unless i am out of options, then it gets expensive for the utility because my sister who is an attourney will find a way to stick them with the entire bill.

Hmm, just curious:

Since you are of the ?? and related to ??, and per the above are in the US on a diplomatic passport, why are you mucking about with a cloud center in a residence ? Did not you know that pot is legal in WA state now, so no need for the absent neighbors to run the grolite operation in their basement with bypassed power ? <G>

?? in the above = site does not support kanji charachters
 
Last edited:
Status
Not open for further replies.
Top