Parelleling generators- why?

[mbrooke]

Is there really any advantage to paralleling emergency generators? I see it all the time on new prints- and heavily pushed by genset manufactures where 4-6 generators will parallel into a single switch board then from that board feed out to the ATSs.


I see several problems with this, including a buss fault in the paralleling switchgear clearing the whole emergency system.


What advantage does this have over say several units which feed individual switch-gear or ATSs? Technically you can easily create several life safety and critical branches or have 3 source ATSs which will provide power even if one generator fails countering the argument of N in paralleling. This design also has lower cost as well in most cases by my estimates.

 

[Sahib]

One reason is economy. A big generator may be under loaded for existing loads wasting fuel. By providing a number of small generators to work in parallel, appropriate number of the generators may be switched on to suit the existing loads and save fuel and SAVE EARTH FROM EXCESSIVE POLLUTION ALSO.

 

[mbrooke]

One reason is economy. A big generator may be under loaded for existing loads wasting fuel. By providing a number of small generators to work in parallel, appropriate number of the generators may be switched on to suit the existing loads and save fuel and SAVE EARTH FROM EXCESSIVE POLLUTION ALSO.

Or save the genset Light loading causes wet stacking- but sizing and load banks can mitigate that.

 

[wbdvt]

Another reason is reliability so you don't have all your eggs in one basket so to speak. Hospitals are a good example where there are multiple generators sharing load. If one generator trips off or doesn't start the other gensets can carry the full load. This also applies when a generator is out of service for maintenance and there is an outage, the other gensets can carry the load.

 

[kwired]

How many generators are "in parallel" that supply the public grid?

 

[Jraef]

How many generators are "in parallel" that supply the public grid?

Ooh, ohh, I know. :bye:
But is there a prize for the correct answer?

 

[oldsparky52]

How many generators are "in parallel" that supply the public grid?

Ooh, ohh, I know. :bye:
But is there a prize for the correct answer?

:lol: You guys are indeed funny. :thumbsup:

 

[JFletcher]

I'd prefer one generator. Less maintenance, less stuff to go wrong.

Last WWTP I worked at had a turbine generator (which was a flaming-out pos) that "upgraded" to two 16 cyl Cats ~ 1MW ea. We had more problems with those than the pos they replaced.

 

[Besoeker]

I'd prefer one generator. Less maintenance, less stuff to go wrong.

Last WWTP I worked at had a turbine generator (which was a flaming-out pos) that "upgraded" to two 16 cyl Cats ~ 1MW ea. We had more problems with those than the pos they replaced.

But a single failure won't put you out of business.

 

[Smart $]

But a single failure won't put you out of business.

I imagine OOB is not often heard at a WWTP. :slaphead:

 

[JFletcher]

But a single failure won't put you out of business.

With a waste water treatment plant, you need the whole plant up 24/7. Sure, most can gravity flow w/o power, and certain loads are more critical than others, but what ends up happening with 2+ generators is one fails and the second can't run the entire plant. Shutting down any part of it usually leads to massive operational problems with DEQ and blown permits.

Having 2+ generators that can run everything by themselves is a cost not even WWTPs will absorb. Not here anyway.

 

[ron]

I prefer to avoid paralleling when I can to avoid custom controllers, but sometimes when it benefits the load, it is easier to implement N+1 generator redundancy with paralleling.

I generally avoid paralleling for redundancy with cascaded transfer switches (or transfer breaker pairs), but it is extra equipment / cost.

 

[Besoeker]

With a waste water treatment plant, you need the whole plant up 24/7. Sure, most can gravity flow w/o power, and certain loads are more critical than others, but what ends up happening with 2+ generators is one fails and the second can't run the entire plant. Shutting down any part of it usually leads to massive operational problems with DEQ and blown permits.

Having 2+ generators that can run everything by themselves is a cost not even WWTPs will absorb. Not here anyway.

Many of the drives I have dealt with/designed have been for the water industry - potable water for the most part. I don't remember any that didn't have spare capacity. And the same for the treatment plants.

 

[mbrooke]

Another reason is reliability so you don't have all your eggs in one basket so to speak. Hospitals are a good example where there are multiple generators sharing load. If one generator trips off or doesn't start the other gensets can carry the full load. This also applies when a generator is out of service for maintenance and there is an outage, the other gensets can carry the load.

You can do the same via a 3 source ATS, correct?

 

[mbrooke]

How many generators are "in parallel" that supply the public grid?

Lots, and often interconnected with something like this:

 

[Sahib]

You can do the same via a 3 source ATS, correct?

What size of each generator will you choose to serve the loads economically at the planning stage? That problem does not arise when you go for paralling of generators because then loads will be shared in proportion automatically by the generators.

 

[mbrooke]

What size of each generator will you choose to serve the loads economically at the planning stage? That problem does not arise when you go for paralling of generators because then loads will be shared in proportion automatically by the generators.

That which will run the life safety and critical branch. If I choose to run everything (including optional loads) I can divide each gen across the optional loads. One gen failing will removed half or 1/3 the optional loads, but because the life safety, critical and equipment are only a small percentage and 3 sourced, transferring to the remaining gen(s) will not require it (them) to be that much over sized. As is I've heard of paralleling applications where all the gens start and they run continuously still loaded under 80% each.

 

[Sahib]

That which will run the life safety and critical branch. If I choose to run everything (including optional loads) I can divide each gen across the optional loads. One gen failing will removed half or 1/3 the optional loads, but because the life safety, critical and equipment are only a small percentage and 3 sourced, transferring to the remaining gen(s) will not require it (them) to be that much over sized. As is I've heard of paralleling applications where all the gens start and they run continuously still loaded under 80% each.

What you propose with one genset down involves extra expenditure including rewiring, downtime, loss of goodwill etc which are avoidable if paralleling scheme is in place. And your proposal fails completely if there is peak shaving scheme also exists which cannot be implemented without automatic paralleling of gensets in your case.

 

[mbrooke]

What you propose with one genset down involves extra expenditure including rewiring, downtime, loss of goodwill etc which are avoidable if paralleling scheme is in place. And your proposal fails completely if there is peak shaving scheme also exists which cannot be implemented without automatic paralleling of gensets in your case.

In an existing building where you can't subdivide the loads.