Unbalanced Voltage

[electau]

The main problem was in the initial design. The PEA should have located the 50kVA transformer near the load. The section of the HV line not on your property is available to other consumers. You would have paid the capital cost but the poles, transformer and conductors would remain the property of the PEA. You would have contributed to the capital cost.

What is your current max demand in amps?

Your reason not to have the HV line extended to near the load is not a logical reason. You would have saved on the cost of conductors and span between poles would be less. Your LV conductors would have been smaller for the same %age voltage drop. It would have been economically cost effective compared to what you have finished up with.

The result that you want is with a given max demand of xx amps per phase you will not have a voltage drop not exceeding 7% (in this case) at any part of your electrical installation.

As yet you have not achieved that requirement.

As an example of the distance problem to supply a load of 10A over a length of 1.6km (1 mile) at single phase at 5% voltage drop would require 70sqmm copper conductor.

To supply a balanced 3 phase load of 10A per phase over the same distance would require 35sqmm copper conductor.

This is why the load supplied must be near the transformer. If the transformer was near your load you could use up to the FL amps per phase, 50kVA is about 70A per phase.

　

[Khonwan]

Thanks to all three of you for your further comments.

Deke, I can confirm that there is no large voltage reduction on startup of the motor. Thanks for your optimism – I’ll take your advice and not worry.

Electau, I saved a lot of cash doing it my way. Your way was going to cost me just short of 1,000,000 baht (I already had PEA survey and quote some 5 years previously) – my way cost just short of 400,000 baht, and that included having to pay half of the third HV cable to extend the 3ph HV 4km to my village. I controlled my costs. I ordered 4 drums of 1,500 meter-long 50mm² and 1 drum of 2,000 metre-long 25mm² aluminium cable direct from the factory (at a substantially cheaper price than retail) some 400km from my home and collected it myself (I owned and was licensed to drive a 6-wheel truck). I spaced my poles (locally produced inferior house-type concrete posts and termite-proof old timber posts purchased from a local temple that was rebuilding) at 40m, similar to the HV pole distances. Two local day-labourers did all the work under my direction.

Hanging the wire was initially a real challenge until we found our technique. I fitted a drum to my Ford tractor, using the rear hydraulics. With block & tackle attached to my 6-wheel trailer plus a system of temporary locks, we were able to stretch the cables very tightly along the whole route with very little height drop. 600,000 baht saved was worth it as far as I was, and am, concerned; my way cost only 2/5th of your way.

You may well wonder why I did not use standard electrical concrete posts: all such concrete posts in the province were made by a company called NPO but PEA told us NPO were no longer supplying since PEA had had a quality issue with them – NPO had just closed their operation (at least in Nakhon Sawan). Neither our local PEA nor their head office in Bangkok could (or would) advise us on where to purchase such posts within or near our province. My wife tried Directory of Enquiries but that drew a blank. We had no access at that time to the Internet. We therefore did the best we could in the circumstances.

I told PEA at the time that I calculated my maximum demand as 17kW (I designed my house to not require air conditioning). Times have changed though, but whatever a new calculation would produce I find I’m able to avoid unacceptable voltage levels (i.e. less than 209V) by trying to balance loads – I monitor this with built-in panel voltmeters on each phase. Part of my strategy is to use 3ph appliances where possible (hence my expressed concern over voltage imbalance). PEA’s adjustment of the taps on my transformer helped enormously.

With respect, I believe you (and all other electrical members on this forum to date from what I’ve seen who provide calculations, Crossy included) are building in a very significant error in your voltage-drop calculations by not recognizing actual voltage as well as nominal voltage. The rated current is based on nominal voltage (220V in Thailand) but actual voltage is often much higher, up to 15% higher. Run your calculation first at the actual on-site voltage (contractually not less than 230V for those with their own transformers) and express the resultant voltage as a percentage of the nominal voltage (220%). I believe your 10A example was based on 220V, which indicated to you that 70mm² would be required – run the calculation again but change the supply voltage to 230V (maintaining the same power rating of 2.2kW, which results in 10A at 220V) and you will find that 50mm² would provide a voltage-drop of less than 5% of 220V. A 9.13% drop on a supply of 230V represents a 5% drop on the nominal voltage of 220V, i.e. both result in 209V. I consider this adjustment to be a correction factor just as necessary as any other correction factors such as temperature or installation method.

By the way, Electau, going back to the welder: I’d mentioned that the manual only refers to the single phase version and that it required 4mm² cable protected by 25A breaker. The actual 3ph machine is plated indicating 20A and the original cable attached to it, supplied by the manufacturer, is 2.5mm². So it seems I’ll not need to change my cable after all. I do understand the reason for this so this is simply for your information.

[Deke]

With respect, I believe you (and all other electrical members on this forum to date from what I’ve seen who provide calculations, Crossy included) are building in a very significant error in your voltage-drop calculations by not recognizing actual voltage as well as nominal voltage. The rated current is based on nominal voltage (220V in Thailand) but actual voltage is often much higher, up to 15% higher. Run your calculation first at the actual on-site voltage (contractually not less than 230V for those with their own transformers) and express the resultant voltage as a percentage of the nominal voltage (220%). I believe your 10A example was based on 220V, which indicated to you that 70mm² would be required – run the calculation again but change the supply voltage to 230V (maintaining the same power rating of 2.2kW, which results in 10A at 220V) and you will find that 50mm² would provide a voltage-drop of less than 5% of 220V. A 9.13% drop on a supply of 230V represents a 5% drop on the nominal voltage of 220V, i.e. both result in 209V. I consider this adjustment to be a correction factor just as necessary as any other correction factors such as temperature or installation method.

For academic interest only, the standard secondary distribution voltage in Thailand is actually 400/230 V, not 380/220 V. It changed, I think, around 2007. So anything built since then should be to the new standard. Also, the voltage tolerances that I mentioned in a previous post on this thread were incorrect. In Thailand they are -10%/+6% at the delivery point (your meter); not the same as the current IEC standard of +/- 10%.

[Khonwan]

Thanks Deke – I was unaware of the change. Some recently purchased appliances are still indicating a rating of 220V.

[Khonwan]

Deke, are you sure about this change?

I was shopping in Big C today and decided to look at the voltage ratings of a number of appliances – I’d say at least half of them were rated at 220V with most of the others 220-230/240V; only a few were specified as 230V.

In consequence, I’ve Googled “nominal voltage Thailand” – I’ve found no sites that state anything but 220/380V, including Wikipedia, Kropla (updated 07-01-10), Telenet (updated 10-04-10), and Wiki-EIG(http://www.electrical-installation.org/wiki/Low_voltage_utility_distribution_networks; updated 11-11-10).

PEA:

http://www.pea.co.th/rtf/Fund%20Webpage/Webpage/Documents/TOR%20Smart%20Meter.pdf is a document written this year by PEA that refers to the nominal voltage as 220/380V. “Guidelines for electrical wiring and equipment installation” http://www.pea.co.th/th/services/services_appendix_c.htm also indicates 220V.

[electau]

One has based calculations on voltage drop as per AS3000 and AS3008.

Voltage drop is based on the nominal supply voltage at point of supply in this case the transformer. Voltage drop calculations based on 230/400V 50Hz 3phase 4 wire supply. Temperature taken as 45C for aerial configuration.

Construction complying with legisative requirements.

You have indicated that your maximum demand was originally 17kW, that is close to 6kW per phase, 26A per phase. It is this value that you calculate your voltage drop. You have given the distance as 1.6km (1 mile).

You have existing 3x50sqmm Al. conductor with 1x 25sqmm Al. conductor for the neutral.

I will calculate the voltage drop figures for the above for copper and aluminium using the tables and fomulae in AS3008.

Edited December 4, 2010 by electau

[Khonwan]

^Thank you, as always, for your reply. Ok, you are compliant with Australian regulations, however, this is not Australia. Australia has nominal voltage of 230V whilst Thailand’s nominal voltage is 220V.

Am I wrong to think that the point of these calculations is in respect to the appliances rather than the transformer? Since nominal voltage here in Thailand is still 220V, appliances made for this market are usually rated 220V. Allowing for a 5% voltage drop (which is less than PEA/MEA’s supply variance), means these appliances are designed to be able to run problem free on 209V. Therefore, isn’t it much simpler, pragmatic and more productive to use the resultant voltage as the end point of the calculation rather than a simple percentage? The way I see it, your way causes people in Thailand to sometimes purchase cable larger, hence more expensive, than is required. Larger sized cable is, of course, only a cost issue but cost is relevant to most people.

By using the Australian 230V for calculation purposes, you are failing circuits here in Thailand that indicate a resultant voltage of 218.4V under load because the drop exceeds 5% of 230V. I cannot see how you can justify this for use in Thailand when 218.4V is actually only a 0.7% drop against the Thai nominal voltage and is therefore completely safe. Your calculations are appropriate for Australia but oppressive for Thailand.

My questions are rhetorical since I don’t think we are going to shift on this matter (though I cannot understand why).

By the way, your statement, “You have existing 3x50sqmm Al. conductor with 1x 25sqmm Al. conductor for the neutral” is incorrect: my neutral conductor matches my phase conductors (50sqmm for first 1500m then 25sqmm on branches thereafter).

[Deke]

I can confirm that 400/230 V, -10%/+6% is the current standard. We just had a project in Chiangmai approved by PEA last week that inclued a 1250 kVA 22 kV/400-230V transformer.

Remember though, that the vast majority of PEA's infrastructure was installed prior to 2007 and therefore conforms to the old standard, not the new one.

[Khonwan]

Thanks for this, Deke. It is all very confusing! But I have to say that my contract with PEA back in 2006 for my 50kVA 22kv/400-230V transformer also specified 230V supply (not 220V). And yet, at least half of the appliances produced by Thai and Japanese manufacturers that I very roughly surveyed in Big C (a store with a high turnover of stock, indicating this was new stock) were rated at 220V, i.e. not even 220-230/240V.

I do agree that transformers are (at least since 2006) designed to output 230V minimum but I don’t see this as conclusive proof that there has been a change in the nominal voltage. It still appears that PEA has not changed their nominal voltage from 220V...or how else do you interpret all the updated website info I presented including that from PEA themselves this year?

My guess is that transformers are being contracted to provide at least 230V in order to offset voltage drop (particularly relative in the case of multi users). Since you have a working relationship with PEA, can you obtain a specific, categorical statement of the present nominal voltage in Thailand?

It is also worth noting that -10% of 230V (207V) is less than -5% of 220V.

[Deke]

You think too much. (I used to hear this from the bar girls 25 years ago).

I'll be meeting with PEA in Chiangmai next Thursday about a project there and will ask them for a standard that I can reference.

[Khonwan]

You think too much. (I used to hear this from the bar girls 25 years ago).

I'll be meeting with PEA in Chiangmai next Thursday about a project there and will ask them for a standard that I can reference.

Thank you.

[electau]

Determination of voltage drop from circuit impedance.

Vd = voltage drop in conductor, in volts.

I = current flowing inconductor, in amps.

Z = impedance of cable in ohms.

For the purposed of this example only the resistance component R has been used.

The actual voltage drop of the conductor is obtined by multplying the cable impedance value by the length of cable and the current as follows:

Vd=ILZ

In circuit applications more tha one conductor is involved and the term L is generally defined as the route length of the circuit.

Single phase. For single phase circuit the impedance of the active( line) and neutral are taken into account. As these conductors are of the same material and the same size, the voltage drop on the circuit is twice what it would be for a single cable.

Vd = 2ILZ.

For a balanced three phase 3 wire circuit no current is flowing in the neutral conductor and at any given instant the current flowing in any one active (line) conductor will be balanced by the currents flowing in the other active (line) conductors. The voltage drop per phase to neutral is the voltage drop in one cable and the voltage drop betwen phases is therefore:

Vd = 1.732 x ILZ

Calculations based on 230/ 400 V 50Hz 3 phase 4 wire supply. -6% to +6%. ie 216.2V to 243.8V. for -6% +10 % this would be 216.2V to 253V. This allows for 240V to be maintained as standard (Australia only).

Thailand is 220/380V -6% +6% which is 206.8V to 233.2V.

We will use 5% for the purposes of this exercise and the 230V/400V values.

Voltage drop single phase 230V is 11.5 volts. Voltage drop for three phase 400V is 20 volts. that is 5% of the nominal voltage.

AC resistance at 50Hz of single core aerial cables with bare or insulated conductors at 45C.

50 sqmm copper 0.438 ohms per km. Aluminium 0.706 ohms per km.

25 sqmm copper 0.822 ohms per km. Aluminium 1.32 ohms per km.

The resistance (impedance) for 1.6km for aluminium conductor is :

For 50sqmm is 1.1296 ohms. and for 25sqmm is 2.112 ohms.

For a single phase 230V load of 10A the voltage drop would be 10 * ( 1.1296+ 2.112) = 32.4 volts. This is 14%.

For a 10A three phase 3 wire load 400V the voltage drop would be 1.732 * 10 * 1.1296 = 19.56 V, this is less than 20V and does not exceed 5%.

For a load of 3.5 amps single phase 230V the voltage drop would be 11.34V.

For 220V values multiply voltage drop figures by 0.9565.

For 240V values multiply voltage drop figures by 1.0434.

If you had a full size neutral 2X25sqmm the single phase voltage drop would be:

10A load 22.5V and for 3.5A load 7.90V.

　

[electau]

I can confirm that 400/230 V, -10%/+6% is the current standard. We just had a project in Chiangmai approved by PEA last week that inclued a 1250 kVA 22 kV/400-230V transformer.

Remember though, that the vast majority of PEA's infrastructure was installed prior to 2007 and therefore conforms to the old standard, not the new one.

Nominal voltage 230/400V -10% + 6% is 207V to 243.8V. The 220/380V -6% to +6% is accomodated within the new range. Caculations will now be made using 230/400V values on new electrical installations in the future.

Thailand has harmonised its standard to meet IEC requirements. Australia did this in 2000. The actual system voltage did not change only the legal definition.

　

[electau]

Transformer no load voltage is determined by the HV voltage and the tapping selected on the HV windings. The no load voltage is usually in practice on the high side of the nominal voltage design rating. So 220V nominal may be 225V in practice.

[electau]

The welder rated at 20A at 230 volts is single phase with 11.5V max voltage drop.

The same welder connected for 400V 2 wire connection is rated at 20/1.732= 11.5A.

This has a voltage drop of 20 volts.

Voltage connection is made within the welder 230V or 400V.

The HWS rated at 12kW is 3 single phase loads of 4 kW each voltage drop 11.5V.

Only the 3 phase motor(s) are balanced loads.