Power Flow Analysis for Industrial Facility 1

[Designer_82]

A large textile mill facility has a problem where the kW's coming in from the utility are double what is being distributed at the loads.
About 1,300 kW's coming in from the utility and about 650 kW's being measured as distributed (via individual branch meters to each utilization branch).
Also their power factor is terrible, about .68
They would like a power flow analysis to see where all the kW's are going.

My thought is gather all cable size/length info from utility transformer all the way to all branches and loads. Have an electrician measure kW's at all utilization branches, downstream & upstream of utility xfmr. Also gather as much specific load info as possible.

Enter all that info in ETAP and run the power flow calculation to try and identify any electrical system inefficiencies.

How would you do it?

 

[sunnysky]

BC Hydro had some suggestions

Here are some of the first things you should check for low p.f.

- Check if you have any rack-mounted film caps rated for xx kVARs
- Large fans and motors only partially loaded have low pf.
- Equipment with a high number of transformers (such as lighting ballasts) will have 10% rated current for good couple but odd harmonic currents increase if voltage is too high.
- Failed or improperly sized capacitors (used for power factor correction)
- Harmonics in a system caused by non-linear electrical loads (e.g. welders)

adding passive kVAR caps are one solution.

 

[jghrist]

First I would check all of the meter settings, PT, and CT ratios and wiring. Even with terrible power factor, 50% losses are not believable. You would have major voltage problems if losses were that high. Do you have any unmetered branches?

Gather all the cable sizes and lengths, transformer ratings and impedances, transformer loss data if available (probably not - use ETAP typical values), motor class and ratings, capacitor ratings, and other load ratings. Then model with ETAP, but use nominal loads, e.g. motor HP, instead of measuring kW at all points. Actual loads may be lower because of underloaded motors, but using full loads would give the worst case losses.

 

[iop95]

Losses of 650kW? A lot of heat... how cool down this?

I think it's an error or multiple errors in measuring.
Check all measuring devices and schematic connections.
Maybe at least a current or voltage transformer has problems.

 

[EnzoAus]

Designer8,

First issue would be to verify the conflicting data. You mention kW but not kWh in your post. Given the variability of power (kW), it is difficult to make comparisons without simultaneous measurements. On the other hand, an energy comparison can be accurately made, if say, monthly consumption figures are available. A large customer should have access to metering data from the utility that would hopefully provide both kWh and kvarh at say 30min intervals so that you can get a very clear understanding of the customer load profile and whether the information seems correct. Knowing for example that a particular large machine of a certain rating starts at a specific time, might provide some verification of the metering.

The next issue is the customer distribution data and it is unclear where the quoted 650kW is being measured. Is there energy metering data that can be summated to compare to the utility metering? A comparison of the two might indicate whether there is a scaling factor (eg. incorrect CT ratio) causing the discrepancy. Don't discount miswiring of CTs leading to incorrect metering.

As others have noted, losses of 50% are not really plausible.

Good luck,

 

[wcaseyharman]

I think I’d start by verifying the metering. Check for proper CT and PT ratios. I’d also see if I could verify with an amp clamp or a PQ meter. You could check with the serving utility - the utility I work for has PQ meters they can install at the transformer.

 

[LionelHutz]

650kW of power just getting "lost" is hard to believe. Either it never existed or it's powering something that didn't get metered.

If they're going to be connecting power metering in a bunch of places then I would suggest just installing permanent meters. There are some models that are cheap enough it's worthwhile. Accuenergy and Trumeter are two that come to mind. They're dirt cheap really for what they do.

If you're doing Etap for such a project when gathering live data, I don't see the point unless you go all in and integrate the metering directly into Etap.

 

[waross]

I encountered that very problem.
I got a free plane ride from Canada to Central America to look into it.
The customer claimed that he was being overcharged.
I checked out all of the metering connections by disconnecting each wire and ringing out each wire.
The wiring was correct but the multiplier had been mis-calculated and was double what it should have been.
The utility gave the customer a credit for about $100,000 dollars based on my findings.

--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[waross]

First check:
With a PF of .68, be sure that the customer is not confusing KVAHrs with KWHrs on the bill.
Be sure that the customer is not confusing KW demand with KWHr consumption.
Put Rogoski coils around the incoming phases and and connect a KWHr meter.
Compare the instantaneous KW with the instantaneous indication on the utility meter.
Please let us know the outcome.


--------------------
Ohm's law
Not just a good idea;
It's the LAW!