Inverter Running at 66% or Transformer

[MJRobinson]

Hello,

Inverter Dilemma.

Is it better to run 21kW on an SMA 30K or

Run 21K on a SMA 20K with a 480Delta to 480/277 WYE transformer?

The Transformer runs about $1,700

Thoughts?

 

[BandGap1.1eV]

Um. What?

You're going to need to outline the situation a bit (much) better than this.

 

[electrofelon]

Pretty vague on the details. Let me take a guess at filling in the gaps. Do you have a delta service? Since the 30k is the only one that will do delta you are wondering about just going with the oversized inverter for simplicity? I see no disadvantage in an oversized inverter. What is the cost difference?

 

[GoldDigger]

Pretty vague on the details. Let me take a guess at filling in the gaps. Do you have a delta service? Since the 30k is the only one that will do delta you are wondering about just going with the oversized inverter for simplicity? I see no disadvantage in an oversized inverter. What is the cost difference?

It seems to me that the STP20000TL-US is in fact a three phase inverter, just with a different output voltage range spec. Which leads to the OP's question of whether that model with a transformer would still be more efficient or more cost effective than a partially loaded 30K model that would not need a transformer.
As you note the 30K can connect either to 480Y/277 or 480 delta, while the lower power models require the 480Y/277 with neutral. I am sure the 30K's initial cost is higher than the 20K.

 

[jaggedben]

From the customer's point of view, the larger inverter would be better since there will be no clipping and no transformer losses. Unless the transformer saved my company a lot of money I'd probably go with the larger inverter. Don't forget to include a bit of extra labor and material, and possibly additional overcurrent protection in the cost for the transfomer. Double check there's no start voltage issue with the larger inverter.

 

[jaggedben]

... I am sure the 30K's initial cost is higher than the 20K.

Yes, but probably in the difference is in the neighborhood of that transformer cost, depending on what pricing they're getting.

 

[ggunn]

From the customer's point of view, the larger inverter would be better since there will be no clipping and no transformer losses.

FWIW and IMO, a 20kW inverter running with 21kW STC DC of modules is unlikely ever to clip.

 

[electrofelon]

To elaborate on what JB said, dont just think of the transformer cost, but also the labor and balance of system components. If it's not an auto, you need a GEC which can be a hassle for a rooftop system. We had to do a bunch of transformers on a rooftop system and it was super time consuming.

 

[jaggedben]

FWIW and IMO, a 20kW inverter running with 21kW STC DC of modules is unlikely ever to clip.

I know, but on a 30kW inverter it definitely never will!

 

[pv_n00b]

From a production point of view it's probably a wash, so look at the relative cost and choose. You can do a simulation of both in SAM but I feel they would come out about the same.

 

[electrofelon]

From a production point of view it's probably a wash, so look at the relative cost and choose. You can do a simulation of both in SAM but I feel they would come out about the same.

Inverter efficiencies at different loadings and DC voltages are a bit of a mystery. I have never seen figures on this. You are saying that PV software models this?

 

[pv_n00b]

Yes, PV software that uses the actual equipment in the project does, not PV Watts. SMA has the Sunny Design software that you can use. It's not financial grade but it will show you any difference you can expect and is pretty simple to use.

Inverter efficiency curves are fairly flat these days across most of the MPPT operating window. They tend to peak at about 80% of full load.

 

[electrofelon]

Yes, PV software that uses the actual equipment in the project does, not PV Watts. SMA has the Sunny Design software that you can use. It's not financial grade but it will show you any difference you can expect and is pretty simple to use.

Inverter efficiency curves are fairly flat these days across most of the MPPT operating window. They tend to peak at about 80% of full load.

Do you know anything off the top of your head about different DC voltages? The fronius primo line of single phase ~240 inverters does 1kv strings, always wondered if you take an efficiency hit for that much bucking.

 

[MJRobinson]

Thanks

Thanks

For one project that is 15kw we are opting for a 20kW with a transformer.

The 21.kW project we are going to use the 30kW. I've used this before on a 19kW system and 11 months later we are still producing what we modeled.

 

[ggunn]

For one project that is 15kw we are opting for a 20kW with a transformer.

The 21.kW project we are going to use the 30kW. I've used this before on a 19kW system and 11 months later we are still producing what we modeled.

Why would you do that? STC conditions rarely if ever exist in the real world. We routinely design and build PV systems with DC:AC ratios of up to 1.2 and they are all monitored. In nearly three years with this company with hundreds of systems I have seen very minor clipping only once or twice. The rest don't clip - ever.

 

[MJRobinson]

Sizing answer

Sizing answer

Why would you do that? STC conditions rarely if ever exist in the real world. We routinely design and build PV systems with DC:AC ratios of up to 1.2 and they are all monitored. In nearly three years with this company with hundreds of systems I have seen very minor clipping only once or twice. The rest don't clip - ever.

We had a SMA20 in stock gathering dust to use with the 15kW system.

The 21kW system we used the SMA30 because of cost of SMA20 + Transformer + Additional Disconnects.

Thanks

 

[Carultch]

To elaborate on what JB said, dont just think of the transformer cost, but also the labor and balance of system components. If it's not an auto, you need a GEC which can be a hassle for a rooftop system. We had to do a bunch of transformers on a rooftop system and it was super time consuming.

How is the external transformer's topology relevant to whether or not a GEC is needed for the PV array on a rooftop? I thought that depends on whether or not there is a transformer in the inverter, rather than an AC-side transformer on the system.

I know GEC's in general are needed for any isolating transformer, but I didn't think that would mean the GEC is required to extend all the way to the PV array.

 

[jaggedben]

How is the external transformer's topology relevant to whether or not a GEC is needed for the PV array on a rooftop? I thought that depends on whether or not there is a transformer in the inverter, rather than an AC-side transformer on the system.

I know GEC's in general are needed for any isolating transformer, but I didn't think that would mean the GEC is required to extend all the way to the PV array.

Maybe if your using the roof for the transformer equipment because there aren't better locations. I can how see a 'CAD monkey' designer with no installation experience might think that's a great idea when it isn't. I think the point is, take the details into consideration. A transformer has to save more money than the cost of the transformer.

 

[pv_n00b]

Do you know anything off the top of your head about different DC voltages? The fronius primo line of single phase ~240 inverters does 1kv strings, always wondered if you take an efficiency hit for that much bucking.

The general rule of thumb is that there is little penalty for bucking and a more significant penalty for boosting. Equipment manufacturers try to avoid boosting if possible and as string voltages go up the need for boosting has gone down. A 480Vac system needs an MPPT voltage of more than 480*√2 = ±679Vdc to avoid a boost circuit, a 277V WYE needs ±392Vdc and a 240Vac system needs ±340Vdc. Just ballpark figures but it gives an idea.

 

[electrofelon]

How is the external transformer's topology relevant to whether or not a GEC is needed for the PV array on a rooftop? I thought that depends on whether or not there is a transformer in the inverter, rather than an AC-side transformer on the system.

I know GEC's in general are needed for any isolating transformer, but I didn't think that would mean the GEC is required to extend all the way to the PV array.

I was thinking of the case where the transformers are on the roof, thus you would need to get a GEC down to the building GES or building steel. On the other hand, an isolating transformer will have taps which may be handy.