Pesky Harmonics / unbalanced residential three phase install

[lplate]

Hi. First post here....

Tomorrow will be an interesting day as I'm borrowing a friendly electrician and his fluke meter to see what's going on with my three phase install.

What we have:

UK based Three-phase four wire PME (Protective multiple earth) supply. 100 amps per phase.

1 phase with pretty much all of the normal 'house' loads (lighting, socket circuits, utility room, immersion coil for the heat store etc.)
1 phase with about 94 amps of convection heaters (nobo as)
1 phase with a 96 amp (several 7Kw elements) tankless water heater which only draws power when someone opens a hot tap.

The 'house' phase has several loads (computers) that I think are likely to cause some harmonics but the trouble's really started now we installed the water heater.....

When it's drawing power you can visably see the house lights pulse. This is especially visible on the 500 watts of dimable lights in the new bathroom if you turn them nearly off.

I'm hoping someone can tell me if this is likely to be simply the cause of my very unbalanced loads being aggrevated even more. Does anyone have suggestions on how I can, cheaply, clean my system up, improve the power factor (save me money), and stop the lights flickering?!

Tomorrow afternoon I should know what the neutrals are doing (hopefully not getting warm), and what, if any, harmonics are present. So I should be able to answer questions.

Finally, can I assume, that the likes of www.liebert.com 's DataWave and www.powerperfector.com products are going to be far too expensive for a SOHO user? I note that most of the manufactures who produce Power Factor correction and harmonic filtering tend to slag off other methods whilst saying their 'patented' method is best. It makes it very difficult to figure out, as a lay person, what's actually 'for real'. Acme's I-Trap and the Harmonix product look like they might do the job too.

Finally I thought you might all get a kick out of a circuit monitoring project I found recently that I plan to implement so I can see exactly what's costing me the most round the house (http://www.kondra.com/circuit/circuit.html)

Cheers,

sk

ps. Hopefully when I've got everything sorted I will add a 15Kw net connected wind turbine

 

[bob]

The 'house' phase has several loads (computers) that I think are likely to cause some harmonics but the trouble's really started now we installed the water heater.....
When it's drawing power you can visibly see the house lights pulse.

You can forget about the harmonics . You don't have enough load to cause a problem. It appears that your load has a high Power Factor so you can forget that also. You did not say anything about yours or the utilities service cable. That could be a source of the voltage drop. The utility transformer could be too far away. Their conductors may be too small for you added load. Suggest you get the utility to do a voltage check.

 

[lplate]

You can forget about the harmonics . You don't have enough load to cause a problem. It appears that you load has a high Power Factor so you can forget that also.

Indeed the figures for both are trivial.

You did not say anything about yours or the utilities service cable. That could be a source of the voltage drop. The utility transformer could be too far away. Their conductors may be too small for you added load. Suggest you get the utility to do a voltage check.

From my perspective the service cables look non-trivial (i.e. tumb thickness) and come off a supply pole just outside the (rural) property. There isn't a transformer there and after going on a short walk I couldn't find one either. I will get on the phone tomorrow to the supplier and see what they have to say. Internally the cable's for the convection heaters are all fresh meeting whatever this years EC directive mandates and star wired (lots of legs) back to the new boards. The water heater is on it's own leg to a seperate isolator and it's own phase. That cable is as large as the service supply cable and shielded as it has to go outside and up the wall to get into the loft.

 

[ramsy]

This is a brave sub-panel project http://www.kondra.com/circuit/circuit3.html for someone with general trade experience.

A few observations / questions:
1) Those sub-pannel feeders should'nt be smaller than service conductors feeding main disconnect? See link to service panel photo: http://www.kondra.com/circuit/images/IMG_0987b.JPG

2) This sub-panel feeder lug for the neutral

appears bigger than its oposite lug in the service panel.


What is going on here? In the service panel, 2nd photo, the neutral feeder looks identical in size to the grounding electrode conductor, perhaps smaller than the phase conductors? Maybe undersized for 100 amps much less for 200.

3) If those photo's were simply distorting my perspective, and feeders were sized and terminated correctly, then my last idea would be switching that new water heater installation over to a phase-to-phase voltage, rather than phase to ground. Higher volts = less amps, and less dimming of the lights.

4) My last remark for your brave project, besides nice looking work, would be perhaps plastic bushings protecting feeders from threaded nipples, and knock-out seals for the holes at the bottom of your service panel, as seen in the above link to service panel photo.

 

[lplate]

This is a brave sub-panel project http://www.kondra.com/circuit/circuit3.html for someone with general trade experience. ... <snip>

3) If those photo's were simply distorting my perspective, and feeders were sized and terminated correctly, then my last idea would be switching that new water heater installation over to a phase-to-phase voltage, rather than phase to ground. Higher volts = less amps, and less dimming of the lights.

Ah... slight confusion. I said 'circuit monitoring project I found', not 'circuit monitoring project I've implemented'! So the feedback on how they did it should really be directed at them.... not me... 8P

Many thanks though as now when I get to implementing this myself I'v got the benefit of your feedback/tips too.

I will have a look at Phase-to-Phase rather than Phase-to-Ground. We hadn't thought about that.

Cheers,

sk

 

[bob]

Iplate
Did you find your transformer? If so how far away?

 

[lplate]

Iplate
Did you find your transformer? If so how far away?

Hi Bob. Sorry for the delay checking back. I'm in the process of launching two new software companies and planning a wedding so hopefully you will excuse any slowness on my part!

Ok. The transformer is indeed not too far away. What fooled me was that the overheads go underground to clear a road but cross OVER it a few hundred yards further up. So, it steps down from two wire medium voltage (?) to three phase and then supplies:

1x light industrial area. Basically a couple of small offices and two small car repair shops (I'm sure they have mig welders and other similar stuff),
30x small homes via single phases and me with three.

Distance from me to the transformer over all the cables must be less than half a mile.

 

[bob]

It should be only 100 ft with the load you are showing and it sounds like its overloaded.

 

[eric stromberg]

What voltage is this system?

 

[lplate]

We're on 240v. In actuallity it floats around 230+- a little.

So to re-cap. Transformers several hundred yards away, when we turn on the 96amp/240v load, the lights on one of the other phases get brighter, from my brief look with a meter there doesn't appear to be any real harmonic issues.

When I finally get hold of the supplier's technical people I would like to be able to tell them what's probably wrong and the reasoning behind it. So...

Why should the transformer be closer?
If it is overloaded why is it pushing more power to the other phases?

Perfectly happy to be pointed to appropriate reading material if that's the best way to answer my questions. I seem to have all the time in the world waiting for the power company to call me back.

sk

 

[eric stromberg]

230/400?

OK, i know i'm going to get flamed for this one, but here goes anyway.
A good rule of thumb is a 'foot per volt.'

A 120 V circuit is good for about 125 feet
A 240 V circuit is good for about 250 feet
A 480 V circuit is good for about 500 feet.

This is just a rule of thumb. Anything above these distances needs to be calculated.

Sounds to me, that, because of the distance, there is a significant voltage drop issue. Let's say that you only have load on one phase. There is voltage drop both directions so the voltage on the phase will be low. If you turn on loads that are on another phase, the amount of current flowing back to the transformer in the neutral will be less. This will decrease the amount of voltage drop on the neutral and will result in the voltage at the load increasing.

Eric

 

[ramsy]

If my review of balanced circuits is correct, opposite phases do cancel current in the neutral. Pardon my ignorance, but checking service-voltage drop this way seems like an elegant idea, since this test may also apply to multiwire branch circuits, and is verified with a simple volt meter and by flipping a few breakers at the panel.

What needs clarification is, excluding other variables, do incandescent lights normally dim when voltage drops below 216vac voltage regulation in EN 50160, or 108vac in the US utilization voltage ANSI C84.

 

[lplate]

If my review of balanced circuits is correct, opposite phases do cancel current in the neutral. Pardon my ignorance, but checking service-voltage drop this way seems like an elegant idea, since this test may also apply to multiwire branch circuits, and is verified with a simple volt meter and by flipping a few breakers at the panel.

What needs clarification is, excluding other variables, do incandescent lights normally dim when voltage drops below 216vac voltage regulation in EN 50160, or 108vac in the US utilization voltage ANSI C84.

Thanks Eric and Ramsy. Your posts really helped me understand what's going on. I'm going to try and do the following over the weekend:

1. Calculate the service cable run distance as accurately as possible (I have a GPS so I should be able to get a pretty good idea). I'm pretty sure it's going to be over 500 feet though.
2. Turn everything off on the house panels except: (a) the lights, (b) the water heater,
3. Measure the voltage change on each phase when I turn the water heater on.

Will post again when I have the results.

 

[bob]

We're on 240v. In actuallity it floats around 230+- a little.
So to re-cap. Transformers several hundred yards away, when we turn on the 96amp/240v load, the lights on one of the other phases get brighter, from my brief look with a meter there doesn't appear to be any real harmonic issues.

The bright lights are a sign of a bad neutral connection. Lights get bright on one leg and dim on the other.

When I finally get hold of the supplier's technical people I would like to be able to tell them what's probably wrong and the reasoning behind it. So...

96 amps is a large load transmit over several hundred yards of conductor. The closer the transformer the less voltage drop. If the transformer is too small you will get a voltage drop also. I think the utility should come out an make a voltage study. Check voltage with no load.
Then begin to add load an see what the voltage does. Take the measurements at you house.
Come back and keep us up to date.