I'm in a project where the specified UPS is a modular UPS with internal N+1 redundancy, the designed drawings SLD shows two parallel UPSs & doesn't show the UPS spec either modular or stand alone, the contractor submitted the two stand alone UPSs for approval, my intention to oblige him for both two parallel modular UPSs with internal N+1 redundancy, if i didn't success with the contractor which of the two is preferred where i'm in a hospital project, one modular UPS with internal N+1 redundancy or two stand alone UPSs which is already N+1 for the whole load?
UPM vs stand alone UPS
- Thread starter m sleem
- Start date
- Status
The internally redundant N+1 UPS would have to allow servicing access where half the unit is powered down and safe for servicing (or faulted) and the other half is hot and running. Guarding and blast barriers come to mind.
I have not seen that type. Does the internally redundant N+1 have two redundant feeders, or one non redundant feeder? Make, model number?
I have not seen that type. Does the internally redundant N+1 have two redundant feeders, or one non redundant feeder? Make, model number?
We have designed two separate UPSs in drawings & as per specs each has to be modular n+1, means that n+1 for the whole load & n+1 for the internal modules.
Well. first question would be the rating at N+1, how does the configuration allow the rating. What is each internal unit rated, then what is allowed at N+1.
Let's say N is 100 kW. Two standalone units at 100 kW each, paralleled, would give you N+1 of 100 kW.
Now, two standalone units, but each internally redundant N+1, dual internal modules. To have an external, per module, 100 kW rating but also internally N+1, each internal unit would also be 100 kW and allow for shutdown and safe servicing of half internally.
There could be several variations. If each internally redundant module is really two, dual 50 kW units, but allow for single unit shutdown and servicing, then two, dual 50 kW units could give you availability of four single 50 kW units and the N+1 rating (three single units) would be 150 kW.
Some clarification of how the rating is achieved would be necessary.
If in the first example, the internally N+1 100 kW units are two 100 kW units paralleled. Two such packages would give you the onsite hardware of four 100 kW units and theoretically, maybe possible, an N+1 rating of 300 kW.
If the above example yields an N+1 rating of only 100 kW because of the manufacturer configuration (half the internal module could not be shutdown for servicing), there would seen to be a lot of excess wasted hardware cost that is otherwise expensive to buy and maintain (matching redundant batteries ???).
As noted, four 50 kW modules would be 150 kW at N+1.
If you have specified two separate UPS rooms, N+1 UPS rooms, and each room is internally N+1, this would be a common topology in a data center. In the DC, each UPS room typically feeds the dual feed fast transfer static switch closer to the load (dual N+1 UPS with dual feed at the load).
Let's say each room, N is capacity to meet the load, is 100 kW. Two rooms, either room can be shutdown totally for servicing. Each room, to be internally N+1 would require two 100 kW units, either may be shutdown planned or unplanned.
This topology for N+1 UPS rooms that are internally N+1, the N capacity at 100 kW would require the onsite hardware of four UPS's and four battery strings, each 100 kW. I believe this is what you have described but it may not be what you have.
Do you have specified dual power at the load?
There would be a substantial hardware savings going to one layer of N+1 and two onsite 100 kW UPS's standalone, to yield 100 kW at N+1 (single feeder from the UPS to the load).
Let's say each room, N is capacity to meet the load, is 100 kW. Two rooms, either room can be shutdown totally for servicing. Each room, to be internally N+1 would require two 100 kW units, either may be shutdown planned or unplanned.
This topology for N+1 UPS rooms that are internally N+1, the N capacity at 100 kW would require the onsite hardware of four UPS's and four battery strings, each 100 kW. I believe this is what you have described but it may not be what you have.
Do you have specified dual power at the load?
There would be a substantial hardware savings going to one layer of N+1 and two onsite 100 kW UPS's standalone, to yield 100 kW at N+1 (single feeder from the UPS to the load).
Please find more clarification, we are in a hospital project & as per the drawings beside the specs, we've two UPSs each 750kva connected to STS (static transfer switch) to achieve n+1 for the whole load which is below 750kva, the specs say each of the UPSs should be modular type may be #4 modules each 200kva means that 600kva n+1 (i would think it was just by mistake, as you explained this connection is more relevant to the DC not hospitals), in case i failed to oblige the contractor to follow the same (since it seems it's not really required for hospitals), which one is better for our case, the one modular ups (#4 modules each 200kva means that 600kva n+1) or the two stand alone UPSs each 750kva?
If your kVA rating is equal I would not know which is better. I would have a guess that the four 750 kVA modules would have a footprint and wiring advantage over eight 250 kVA modules. The other factor in my mind would be preferred brand name and servicing arrangement. This is assuming the modules are equal in other areas (both fully online double conversion type?). Really, you would have to value engineer it, price it both ways to see whats the cost difference is.
Just to be clear what you have specified, two UPS's rated 750 kVA (600 kW) which are dual feed to the static switch for N+1 at the load. Each UPS to be internally N+1 is two paralleled 750 kVA units (four onsite total) or four paralleled 250 kVA units (not 200) (eight onsite total).
I would assume there's a manufacturer difference between the packages as one manufacturer would not normally offer it both ways. The manufacturer would have a preference for four 750's vs eight 250's.
Is this correct, is this your choice? There are a lot of other factors to consider. You could be doubling the breaker count at the input/output switchgear.
I hope you understand me correctly, why four? why 250 kva?Anyway we've a three options descending from the optimum to down as follows:
- Two 750kva modular UPSs each 3+1 of internal UPM each 200kva (in this case we cannot request 250kva from the contractor).
- Two 750kva stand alone UPSs (doesn't care about the no & rating of the internal modules).
- One 750kva modular UPS 3+1 of internal UPM each 200kva (in this case we can request 250kva rather than the 200kva).
Of course, the first is the best, just in case we cannot proceed with the first, which of the other two is better?
Thanks in advance.
The question for clarification is, from the UPS to the load, are you running a single feed to the load or dual feed to a fast transfer static switch, then the load.
option 1 describes dual feed to the STS then load, onsite total installed hardware of 2( 3+1) or eight 200 kVA units.
option 2 describes dual feed to the STS then load, each UPS not internally N+1 redundant, onsite total installed hardware of 2(1) or 2 750 kVA units.
option 3 describes single feed to the load, no dual feed to the STS, single UPS internally N+1, onsite total installed hardware of 1(3+1) or four 200 kVA units.
There's an inconsistency in the descriptions of, are you using dual feed to the STS in the topology, and if the individual UPS sources are internally N+1. The options are not described as an apples to apples topology.
- Status