Transformer sizing

[mehdi1351]

I would like to know whether I should size my Transformer based on "Connected loads" or "Operating loads".
I know the first one is a bit more conservative but I want to know what is the the standard practise

Thanks

 

[jim dungar]

I would like to know whether I should size my Transformer based on "Connected loads" or "Operating loads".
I know the first one is a bit more conservative but I want to know what is the the standard practise

Thanks

The NEC provides no guidance as to how to size a transformer.

I have always used 'operating loads'.

 

[mehdi1351]

Yap, I am thinking like you. It is more sense to size the transformer based on operating loads rather than connected loads but i was not sure about american standards, NPFA70 etc.

 

[Jpflex]

I would like to know whether I should size my Transformer based on "Connected loads" or "Operating loads".
I know the first one is a bit more conservative but I want to know what is the the standard practise

Thanks

My boss used to order a transformer but did not know how to do load calculations and would constantly add new loads after purchase so he would often h ave to buy another larger transformer

 

[kingpb]

I would like to know whether I should size my Transformer based on "Connected loads" or "Operating loads".
I know the first one is a bit more conservative but I want to know what is the the standard practise

Thanks

How are you sizing your panel that has the loads connected to it? Simply pick the closest (rounding up) size to the panel. Some will size the transformer for the full capacity of the panel or board, it is connected to if there is a very likely chance that additional loads will need to be added later.

 

[KevinKos]

The standard practice for transformer sizing generally involves using "Connected Loads" rather than "Operating Loads." The connected load represents the total possible load that could be connected to the transformer, providing a more conservative and safer approach, especially for sizing the transformer to handle peak loads and future expansion.

Key Points:​

1. Conservative Approach:
   • Connected Loads include all the possible loads that might be connected to the system. This ensures that the transformer can handle the maximum demand without being overloaded, providing a margin of safety and allowing for future expansion without the need for immediate upgrades.
2. Practical Considerations:
   • Operating Loads reflect the actual load expected under normal operating conditions. While this might be closer to real-world usage, it doesn't account for peak loads or potential increases in demand over time.
3. NEC Guidelines:
   • According to NEC guidelines, transformer sizing often considers the maximum possible load to ensure safety and reliability. For example, NEC Article 450.3 specifies overcurrent protection based on the maximum current expected under peak load conditions .
4. Efficiency and Cost:
   • Using connected loads for transformer sizing might be more expensive initially due to the higher capacity required. However, it avoids potential costs and downtime associated with upgrading undersized transformers later.
5. Standard Practices:
   • Electrical utilities and industry standards typically recommend using the connected load to size transformers to avoid overloading and ensure the transformer can handle unexpected surges and future load increases.

Conclusion:​

Sizing transformers based on connected loads is a standard and recommended practice due to its conservative nature, ensuring safety, reliability, and accommodation for future load growth. While it may seem more conservative and potentially costlier initially, it provides a robust solution to avoid the risks and costs associated with transformer overloads and upgrades.

For more detailed guidelines and specific calculations, refer to NEC Article 450 and other relevant sections that discuss transformer sizing and protection .

 

[jim dungar]

For the load to operate it must be connected.
But, if it is connected it may not be operating. Not all connected loads are coincident.

The NEC contains no language about the sizing of transformers.

Consulting engineers routinely oversize transformers in an effort to cover themselves under the guise of unexpected future growth. Isn't their it job to anticipate the future?
Utilities like to undersize transformers selected using NEC load estimates in an effort to reduce transformer losses.

Some of the early DOE transformer requirements were to base efficiencies at 35% loading as their surveys, in the 80's, showed this was a typical average loading in offices and schools.

 

[mehdi1351]

@KevinKos and @jim dungar thanks for the response. I guess there should not be a prescription on this matter. I think it is thoroughly depends on the type of the loads, i.e some connected loads are duty/standby and it goes without saying that both duty and standby won't come into the line regardless of whether you are going to increase the plant capacity or not.
In Middle East, we size the transformer based on IEC60076, which is based on "operating loads" plus 35% over specially due to temp effect on the transformer de-rating and then we do the adequacy checks to see the selected KVA would meet the requirement.
BUT as I want to manufacture a new plant in Texas, I don't really know what is the US regulations, i.e NEC, NFPA70 etc tells about the Transformer sizing.
If you really sure about that those codes & Standards telling about transformer sizing criteria, please share it with me
Thanks heaps
Mehdi

 

[Knightryder12]

I always size our transformers on connected load. If the connected load falls in between standard size, then we go up to the next standard size transformer.

 

[jim dungar]

I always size our transformers on connected load. If the connected load falls in between standard size, then we go up to the next standard size transformer.

Do you assume all connected loads are all coincident. Do you ever consider ongoing costs of ownership of unused transformer capacity?

 

[Knightryder12]

Do you assume all connected loads are all coincident. Do you ever consider ongoing costs of ownership of unused transformer capacity?

As far as I know, there is NO exception for non-coincident loads in the general sense. Non-coincidental loads are only allowed for the likes of mechanical equipment, etc. I guess if what the transformer was feeding owner provided equipment that the owner has stated would not be operated at the same time I would figure that in. But from what my experiences have been with submitting plans to the AHJ, they would need a letter from the owner stating that that equipment would never run at the same time. That also pertains to mechanical equipment, the controls would have to state that those pieces of equipment cannot and will not operate at the same time.

 

[jim dungar]

I am not saying you are wrong. But, the NEC does not address the sizing of the transformer itself, even though there is guidance on overcurrent protection and conductor sizing based on your connected loads.

 

[petersonra]

I suspect in most cases if you size a transformer of any significant size based on the nec calculated load you will be oversizing it by a lot. I would make an effort to come up with a reasonable actual load. It's not hard to do.

I do something similar for sizing cabinet air conditioners. If you look at VFD heat dissipation charts, it's not unusual for a load at 75% of the max load a vfd can put out will result in about a 50% production of the Max heat it would produce at full load. And vfds almost never run at full load.

It's not real hard to come up with a realistic load if you put some effort into it.

I have seen mccs sized based on max loads running 100% of the time, when half the motors are sump pumps that run for a few minutes at a time every few hours. Or chemical pumps that run for half an hour twice a week.

 

[jim dungar]

I suspect in most cases if you size a transformer of any significant size based on the nec calculated load you will be oversizing it by a lot. I would make an effort to come up with a reasonable actual load. It's not hard to do.

I do something similar for sizing cabinet air conditioners. If you look at VFD heat dissipation charts, it's not unusual for a load at 75% of the max load a vfd can put out will result in about a 50% production of the Max heat it would produce at full load. And vfds almost never run at full load.

It's not real hard to come up with a realistic load if you put some effort into it.

I have seen mccs sized based on max loads running 100% of the time, when half the motors are sump pumps that run for a few minutes at a time every few hours. Or chemical pumps that run for half an hour twice a week.

These oversizing situations are reasons why some people use transformers with cooling provisions that allow the transformer to be overloaded, such as fan cooling or higher temp allowances.

 

[kwired]

As far as I know, there is NO exception for non-coincident loads in the general sense. Non-coincidental loads are only allowed for the likes of mechanical equipment, etc. I guess if what the transformer was feeding owner provided equipment that the owner has stated would not be operated at the same time I would figure that in. But from what my experiences have been with submitting plans to the AHJ, they would need a letter from the owner stating that that equipment would never run at the same time. That also pertains to mechanical equipment, the controls would have to state that those pieces of equipment cannot and will not operate at the same time.

Most situations you wouldn't heat and cool a particular space at the same time so you use which ever mode is the higher load for sizing supply items.

Maybe you have an industrial process that has say production level loading and then cleaning level loading. Some the cleaning loads may never run at same time as some the production loads, you certainly wouldn't need to design the supply for this to handle all items if they would never run at the same time.

You also sometimes look at demand data when determining whether you have enough capacity to add new loads, in many instances it becomes obvious the actual demand is well below what you would provide based on "connected load".

 

[Knightryder12]

Most situations you wouldn't heat and cool a particular space at the same time so you use which ever mode is the higher load for sizing supply items.

Maybe you have an industrial process that has say production level loading and then cleaning level loading. Some the cleaning loads may never run at same time as some the production loads, you certainly wouldn't need to design the supply for this to handle all items if they would never run at the same time.

You also sometimes look at demand data when determining whether you have enough capacity to add new loads, in many instances it becomes obvious the actual demand is well below what you would provide based on "connected load".

The heating and cooling thing I get and that is in the NEC. But we get no other major diversity in the NEC. Everything we do is over sized. I know pretty much every service I design is usually way over sized.
I do a lot of pharmaceutical work where the heat is used for dehumidification of the space, so the heat and cool will run at the same time. But typically I am not feeding that from a dry type transformer.

I dare you to submit a set of calculations for permit where you have, lets say, 47KVA connected load on a 45Kva transformer. AHJ will reject that.

 

[petersonra]

The heating and cooling thing I get and that is in the NEC. But we get no other major diversity in the NEC. Everything we do is over sized. I know pretty much every service I design is usually way over sized.
I do a lot of pharmaceutical work where the heat is used for dehumidification of the space, so the heat and cool will run at the same time. But typically I am not feeding that from a dry type transformer.

I dare you to submit a set of calculations for permit where you have, lets say, 47KVA connected load on a 45Kva transformer. AHJ will reject that.

On what basis would they reject it? There is no code violation.

 

[Knightryder12]

On what basis would they reject it? There is no code violation.

You are overloading a dry type transformer. Those do not like constant overloads. Been dinged on it before.

 

[jim dungar]

You are overloading a dry type transformer. Those do not like constant overloads. Been dinged on it before.

Totaling connected loads does not provide evidence that a transformer is being overloaded, but totalling operating load does.

If the transformer is properly protected, per 450.3, can it really be overloaded?

 

[Knightryder12]

Totaling connected loads does not provide evidence that a transformer is being overloaded, but totalling operating load does.

If the transformer is properly protected, per 450.3, can it really be overloaded?

Isn't that what the NEC has us do? By your logic, its ok to have 160a connected load on a 150a rated panel. That does not make sense. Again, if i can prove that certain loads are no coincidental, I will use that to my and the owners advantage. But as I have said before, show me in the code that allows that other than heat/cool and receptacle load.
I have been doing design for a long time and have been trained by some very knowledgeable Electrical engineers. I do not know of any other diversities we can take on loading/service/panel/transformer sizing other than the ones I mentioned.
I have had the owner provide a letter to the AHJ to state that certain equipment will never be used at the same time and have decent luck to have the AHJ accept that. But not all will.

If there are more exceptions in the code, please let me know. i am always up for learning. Also, how much of a loss are talking about between a 45 and 75kva transformer? I have never done the math, but I would presume its negligible.