Off Grid Building 2

[Mbrooke]

So I'm designing a co-gen power system for an off grid remote research building. I'm not really sure where to begin- I need a few pointers/guidance on where to start...

Does this look like a good design? Should I get creative with how the gensets are interconnected?

Or just keep things really basic?

Anything I've missed? Go/no go? Or am I way off?


Power is normally produced by natural gas and solar with onsite diesel tanks to meet the requirements of article 700.

https://files.engineering.com/getfi...abd-4b3c-8a21-98610d4667c5&file=IMG_0040x.jpg

https://files.engineering.com/getfi...ad1-4634-8dbf-62914832e2a8&file=IMG_0041x.jpg

 

[cranky108]

Just my 2 cents, but if you are wanting reliability I would go with a breaker and a half bus arrangement.
That maybe harder to do with lower voltages. Or maybe not cost effective.

 

[Mbrooke]

I've never seen breaker and a half used on LV, nor have I thought about it... it would make for some interesting operation and relaying though.

Ultimately I don't want to be running these gens in parallel, hence the ATS approach seen here.

 

[catserveng]

Why do you not want to operate the units in parallel? Typically mixed units plants operating in island mode are all working thru a paralleling arrangement of some kind for lots of reasons.

One of your biggest problems will be restoring after an outage and bringing the plant back up on diesel, then trying to transition over to the gas engine driven units. Been down that path several times, I think you will find your end user won't like it.

Woodward has added solar integration to the EasyGen line of controls, you may want to talk with them as it is likely a solution that best fits your plant and generator sizing.

Attached is a product guide to the new Solar integration.

MikeL.

 

[Mbrooke]

Paralleling adds cost, complexity, and common failure modes that I would like to avoid. Most can service an ATS during an emergency, but if the gear fails in particular the logic good luck.

The gas units have automatic electric starters that get the engine up to speed (and voltage) within 30 seconds. Some models are listed as capable of accepting full load in one step.

 

[davidbeach]

The ATSs are the weakest link, single point of failure. Breaker based paralleling gear can be designed to be more reliable. What's an hour of no power to part or all of the facility worth? Is not paralleling really cheaper?

I’ll see your silver lining and raise you two black clouds. - Protection Operations

 

[Mbrooke]

Can't ATSs be designed more reliable? Gear with computers seems far more complex with more parts to them. I've never understood that one.

 

[Mbrooke]

@David: I notice you've brought up ATSs before. With all the well earned experience on your tool belt, I've take it you've seen a fair share of ATSs failures?

 

[FreddyNurk]

I agree with catserveng, if this is really an islanded application with no grid connection, make it capable of running in parallel to start with.
As catserveng has mentioned, Woodward has a decent range of products to support this, so does ComAp.

Paralleling allows:
1. Choice of the most cost effective generation for the load at the time, without the penalty of having to go black to change generation sources.
2. Transparent transition between generation sources without affecting consumers.

There is the consideration of equipment ratings for the required switchboards, but its not insurmountable and probably no worse than multiple ATSs to feed different sections of plant.


EDMS Australia

 

[Mbrooke]

How much extra are you spending on paralleling gear?

I just can't see a compelling need for it.

 

[Mbrooke]

Ok, so I came up with a bit of a compromise.

Size both gas gens to 100% of the switch gear load (general lighting and power + emergency circuits).

Combine feeders under 400 amps.

Include a tie between both switch gears so that either gas gen can be serviced during light load periods ie when cooling is not needed.

Still got ATS in series, though why not?

 

[xnuke]

You can feed from either side with the 1000 kW generators by opening one generator breaker and closing the tie (assuming an open transition is okay) using electrically-operated breakers, essentially created a breaker-based transfer scheme, so you don't need the first set of ATSs.

Also, please reduce the clutter by breaking this up into several diagrams. There's way too much on this drawing.

xnuke
"Live and act within the limit of your knowledge and keep expanding it to the limit of your life." Ayn Rand, Atlas Shrugged.
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[itsmoked]

I know the feeling Mbrooke. I like to see everything in my mind which multiple pages screws-over. My old boss insisted you always use "17" x 24" artist pads or you will cramp your thinking". He was a prodigious inventor of the highest order.

Alas, I can only get about 11 x 17 to work for me.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[Mbrooke]

@Xnuke: I like your idea, but the PLC will also need to shed the chiller mains (and perhaps some load) on one side. The 1000kw rating is 100% of the normal load on one side of the gear.

Originally I was thinking 50% loading and more gens.


Here is "zoomed in" version without all the extras.

 

[Mbrooke]

@Its smoked: Thanks Multiple pages are like a train derailment for me, I lose track of the bigger picture. At the same time I've gotten used to re-drawing it all on a single sheet of paper so I can make sense of it all. Sometimes I forget that not everyone is like me.


Anyway, to be fair, the drawings are pretty bad by my standards.

Right now this project is all napkin drawings, trying to find what "looks good".

 

[FreddyNurk]

At least some of the more modern equipment that handles load sharing and synchronising can also handle load shedding of certain loads without the need to resort to a PLC and all the customisation that goes along with it.

Another factor, sizing of generators has a number of requirements, often competing, in terms of what your operation can handle. Paralleling of sets makes for much easier starting of larger loads, and once running excess sets can be dropped off without any penalty. Attempting to do this with what I presume are open-transition ATSs would likely result in the need to oversize the sets to be able to start large loads, including the associated fuel efficiency penalties.

EDMS Australia

 

[Mbrooke]

Yes, open transition ATSs. You make a good point, the generators do have to take inrush and motors into account.

The heat from the gas units has a high probability of being reclaimed from the engine jacket and exhaust.

However I'm figuring out you get the most bang from fuel when the generator is 60-90% loaded. so 50/50 loading will probably be abandoned.

 

[itsmoked]

You could seriously mitigate big load starting with soft starters or VFDs even if they have a slight efficiency loss as compared to added more generators or having to oversize them (large efficiency loss).

Keith Cress
kcress -

http://www.flaminsystems.com

 

[FreddyNurk]

itsmoked, that's true for loads that can be started by a VFD or softstarter. Probably not so relevant for dumping the load onto a set via an ATS, at least not without a complicated staging arrangement to get everything back on.
However, in this case the money spent on such equipment could just as easily cover the cost of decent Generator equipment and motorised breakers which allow for paralleling, and whilst it won't solve all the cold load pickup issues, it'll certainly be better equipped to handle it than a bunch of open transition ATSs.

MBrooke, have you actually asked the client how they feel about having to go black every time there's an issue, a set needs to be serviced or they need to change sets?
I've never met a client that is happy about having to turn the power off if there are other viable options.

 

[Hoxton123]

Carrying on from FreddyNurk's comment above, whats is your estimated genset availability? Say its 90%, 8760 hours in a year, so the set could be down for 876 hours, that's 36.5 days!

95% gives you 18 days and so on...

Admittedly, some of this will be planned maintenance (oil changes, tappets etc) which could be scheduled at weekends - if this site does not operate at weekends. Breakdowns can occur randomly.

Next, what is the load profile? The gas engines I worked with would shut down if the load fell below 30% for more than a few minutes, and this could not be overridden.

I would just run them in parallel with N + 1 or N + 2,

Parallelling this type of set is a mature technology, changing from ATS to parallel later will not be easy!

I wold look at the load profile to determine the rating of the sets.