The first approximation is that the difference in ratios is the difference in size.
Eg: A 10% step-up, the auto-transformer will be 10% of the size of the isolating transformer.
For a step down the ratio is a little different compared to a step up.
Depends on three transformer solution or a two transformer open-delta solution.
Depends on the nearest common transformer size.
For a large change in voltage relative to the supply voltage;- Not much advantage.
For a small change in voltage relative to the supply voltage;- A good advantage.
Common adjustments where auto-transformers are frequently used:
208:240 Volts.
480:600 Volts.
I have seen where site conditions required a voltage adjustment from 480 Volts to 600 Volts and from 600 Volts to 480 Volts.
The adjustment was done with auto-transformers.
Then in both cases I have seen small equipment that required the original voltage.
The most economical solution was a second set of auto-transformers going back to the original voltage.
That's right; 480 Volts to 600 Volts and back to 480 Volts.
And 600 Volts to 480 Volts and back to 600 Volts.
I believe that this speaks to the robustness and efficiency of auto-transformer solutions.
Bill
--------------------
"Why not the best?"
Jimmy Carter
Sorry. That's out of my league. One issue with all step down auto-transformers is the possibility of the higher voltage being impressed on the lower voltage circuit in the event of an open primary.
Bill
--------------------
"Why not the best?"
Jimmy Carter
I think AT changes are reasonably limited to the participating insulation systems. Boosting or dropping from 208V to 240V or possibly even 480 to 600V usually keeps everything in the same class such as 300V or 600V insulation systems.
Reaching out of the same insulation systems like 500kV down to 138kV would open the lower insulation system to catastrophic system failure (likely wide scale destruction) in an opening winding fault. None of the lowside system protection would even work successfully.
"Reaching out of the same insulation systems like 500kV down to 138kV would open the lower insulation system to catastrophic system failure (likely wide scale destruction) in an opening winding fault. None of the lowside system protection would even work successfully."
I would disagree- you could program relaying to pick this up- something as simple as negative sequence or neutral over current could pick it up. Though you do bring up a good point on how the system would tolerate it in the mean time before the breakers clear. But then again open winding faults are less common in such large power transformers. Insulation break down is usually what does them in- aside from tap changer faults.
How do they vary in size and weight relative to MVA? The size and weight of an auto transformer compared to a two winding transformer will be that of a two winding transformer of MVA of (HV-LV/HV)x line MVA where HV.LV are the voltages of HV &LV circuits. This factor is called co-ratio of auto connection. 100 MVA 220/132 kV auto will have losses, weight, size and cost of a 100x(220-132/220) =40 MVA two winding transformer. Losses, weight and cost of a 40 MVA two winding transformer will be (40/100) raised to 0.75 of a 100 MVA two winding transformer. It means only 40 MVA is transferred through the transformer action and balance 60 MVA is transferred directly to secondary through the galvanic connection. This is true irrespective voltage class, LV or EHV. In reality, weight, size and price may be more for auto than above formula because of tap changers, tertiary etc. Usually auto connection is economical up to a voltage ratio of 1:3 in case of EHV and 1:4 in case of LV voltages. So you will see 400/132 or 110 kV auto transformers and not 400/66 kV. Technically 400/66 KV auto is quite feasible but it will be costlier than two winding units due to the need of 3 nos line end tap changers. So lower the voltage difference between HV and LV voltages, more economical will be the auto connection.
Some history of auto-transformers. Auto connection was discovered and experimented by engineers during the initial years of transformer engineering in 19th century. But the breakthrough and huge demand came in the first decade of 20th century. By then 110 or 220 V distribution voltage became common in Europe. Electricity was used mainly for household lighting using carbon filament lamps. Then the great breakthrough came by the invention of metal filament lamps. Many metals were tried and tungsten filament was the winner by OSRAM lamps. But the problem was engineers could not make thin filament then and hence resistance of lamps was so low that 50 V AC had to be used. Solution was to provide 200/50 V single phase air cooled auto-transformer in each house. Still it was attractive solution as we saw with LED lamps recently. In 1905,in London,UK electricity was billed at 4.5d(pence?)per kWH. Cost of a 16 cp(candle power)carbon lamp was one shilling ( but consumed 60 W) while cost of 16 cp metal filament lamp was 3 shillings ( but consumed only 20 W) A 1KW ( KVA came much later) 200/50 V auto transformer cost was 2 pounds and 19 shillings. Losses in a 0.5 kW auto transformer was 10/24 W with a full load efficiency of 93.6 %.
Later when transmission voltages doubled with each decade (11,22, 66,132,220 kV ) the ideal transformers for interconnection were all auto transformers as voltage ratio of 2 gave good economy. During initial years problem was the non availability of high voltage line end on load-tap changers and hence various costly alternate solutions had to be used. But by the middle of 1960's line end OLTC appeared in the scene making auto-transformers more attractive.
@PRC-Thank you! Thats the equation I was looking for.
"Technically 400/66 KV auto is quite feasible but it will be costlier than two winding units due to the need of 3 nos line end tap changers. So lower the voltage difference between HV and LV voltages, more economical will be the auto connection."
Assuming you do not have tap changers- an auto would still be economical here albeit not the same extent (83.5)? Tertiary- a rough multiplier?
Thank you Edison.
Mbrooke, the effect of tertiary is about 7 % increase in weight /size. But why do you want a tertiary. If the unit planned is of 3 phase 3 limbed construction, you can avoid it without any major effect. In India there are many 100-160 MVA 220/132 kV auto-transformers in service with out a stabilizing tertiary winding.
On a side note, you mentioned lamp filaments. That is an area where much improvement has been made.
In the 40s the large diesel engines needed a 24 Volt starting system for reliable starting. A 24 Volt lamp filament is roughly twice as long and 1/4 the cross section of an equivalent 12 Volt filament.
The available 24 Volt lamps would not stand the shocks and vibration of freight truck service, and the series parallel systems where developed.
The trucks where built with a basic 12 Volt electrical system. A second battery was connected and charged in parallel.
When starting, the second battery was connected in series by the Series Parallel Switch and 24 Volts was supplied to the starter only.
These days roads are better, suspension systems are much better, batteries are better, starting systems are better and lamp filaments are much more robust.
These days many trucks use a 12 Volt system, possibly with two batteries in parallel.
24 Volt systems with 24 Volt lamps are common in construction equipment such as dozers and excavators.
I haven't seen a series parallel switch on a newer truck for years now.
And another instance of filament strength was street lighting on street car routes. In one city the vibration of the rail cars led to frequent failure of the street light filaments.
The solution was to use 20 Amp series lamps. The filament in a 250 Watt, 20 amp lamp was so massive that it didn't go out quickly, it cooled down like a toaster element.
To avoid the relatively high losses and cost of copper associated with a 20 Amp series circuit these lamps were typically supplied by a 7.5 Amp series circuit. Inside each Lamp standard was a small 7.5 A to 20 A current transformer.
That problem was solved when the street cars were phased out of service.
Bill
--------------------
"Why not the best?"
Jimmy Carter
Mbrooke, the effect of tertiary is about 7 % increase in weight /size. But why do you want a tertiary. If the unit planned is of 3 phase 3 limbed construction, you can avoid it without any major effect. In India there are many 100-160 MVA 220/132 kV auto-transformers in service with out a stabilizing tertiary winding.
Perhaps another thread, however I have always seen a tertiary on the auto-transformers around here even when its a 3 phase unit. Its just something thats done by default.
From an engineering paper published in India. I have never heard this, but they say that 3:1 is not used because of some type of disturbance. Anyone know what they are referring to (see attachment)?
To my understanding that info is not correct.You may get the saving % mentioned in losses but may not in prices esp if EHV and OLTC is involved. Price may be same or only small savings. The limit of 1:3 in voltage ratio is only due to economical reasons as you exceed that ratio the savings become negligible compared to a two winding transformer - the % of power directly moving to secondary due to galvanic action comes down with higher voltage ratio.
In fact, some disturbance may come if the ratio is small and not when large. When a LG fault occurs on secondary terminal or an impulse surge comes on primary terminal the entire primary voltage or coming surge will be seen by the small series winding. Transformer engineers have solutions to deal such situations.
I am sure its not far off for HV/EHV.