1MW system design

[JTSOLAR]

Hey all,

I've been given the task of pricing out a 1MW solar project. I don't have much of a background in solar system design but I'm trying to understand how a typical system of this size is constructed and what equpiment is needed in order to come up with an accurate cost for a 1MW project. I have a few questions and I was hoping you guys could help me out.

The panels we will be using for this system have a Voc of 291.4V, Isc of 2.75A and a power output of 530.5 Watts. Maximum system voltage is 1000V.

What factors should I look at when connecting these panels into strings/arrays? I'm thinkng of putting 2 panels in series to give me a string voltage of 582.8V and then putting six strings in series to go to my combiner (with a 20A rating) with a current of 16.5A.

What equipment is needed to conect the output from the inverters? Do they just connect to the inverter?

Do I need a dc dissconnect after each combiner?

Any input you guys have would be helpful,

Thanks,
JT

[Mike90250]

Wow ! Awesome ! And so far out of my residential league !

Generally, with the large installs, you use 3 phase inverter systems, and lots of
planning with both the city and the utility company.

That's all I have.

[JTSOLAR]

Haha..thanks. It's very much in the planning stages. I'm just curious to figure out how it works so I can figure out how much it will cost.

[Aussie Bob]

JT,

Are you an Aussie by any chance?

[JTSOLAR]

Close, Canadian.

[JTSOLAR]

Here's the issues I'm currently scratching my head about...How many combiners do I need? Say I connect 2 panels in series, then I connect three sets of two in parallel for a total of 6 panels on a string. Would I then connect two strings together at a combiner? Or would it make more sense to connect more then two at a combiner? What is considered when combining strings at a combiner? Also, can I connect the outputs from the combiners directly to the inverter or is there another piece of equipment that gets connected in between?

Any ideas?

[Mike90250]

You want to set up series strings to your max (Cold Morning Startup voltage) input.

Each string is paralleled in the combiner box. You use as many strings as needed to "fill
your inverter" so you don't have to buy so many inverters. The combiner box should be
spec'ed for X number of strings. The inverter should have a spec for how many input wires it can accept. Those feed wires come from your combiners. If 3 feed wires, you need 3 combiners. Then you figure what size combiners you need , 3 or 12 circuit.
Each string, in the combiner, should have it's own fuse or breaker.

[Photovoltaic Supporter]

Hello All, Wow that's a nice project 1MW system, might I ask if it's a local company that's asking or you simply doing this for someone else...to submit on their behalf? I ask because just the design work involved is a fare part of the cost factor so I am wondering how you might do the end cost pricing without all the figures in mind? I'd say you need to simply do the math at about 8-9 per watt and that should cover it...I think, if anyone thinks I'm way off please let me know. I too am having more trouble pricing out projects then anything else involved. So I've started using simple watt times dollar amount to get end result cost of project. Seeing there are so many different models, manufactures and add ons...it's hard to do prior to being awarded the job. You should keep in mind that an over ride should be in order here...of at least 15% or so!

[JTSOLAR]

Yes, the company I work for is planning on funding the project. They have asked me to look at the project and see if it would be viable. They have a particular panel they want to use (see specs in the first post) but everything else needs to be priced out. I'd love to go to them and say it'll cost $8 bucks a watt but they will want to see the break down. So in the mean time I'm trying to learn as much as I can about how this type of project is desgned and put together.

[MarineLiner]

the task of pricing out a 1MW solar project.
The panel:
Voc of 291.4V,
Isc of 2.75A and
power output of 530.5 Watts.

How is it getting on?
1MW = 1,000KW = 1,000,000W
Need at least 1,886 panels.
Space for panels only: 5,000m2 at least. Assuming every 1m2 panel produce 200w.
Thought, it is not the "on roof system".

[newenergy]

What module is this?

Seems like you need to select your inverter(s) before you know what the design will look like in between the modules and the inverter(s).

[JTSOLAR]

It's an Applied Material SunFab module.

[astrayan]

A lot of the cost of a solar install seems to be to do with the mounting, and whether it tracks.

These 40kW inverters look amusing, and most seem to be around $1 per watt.

http://www.energymatters.com.au/fronius-40000w-40kw-grid-connect-inverter-p-917.html

I'm guessing that a major issue would be how 25 of these inverters would distort the local grid waveform. They would all have to be synchronized, or it would get bumpy.


Then there's the matter of the transformers, necessary to get the voltage up to the multi-kiloVolt distribution voltage.

[mountain]

Converting from a DC bank to 600V 3-ph, at 1MW, is possible with off-the converter shelf equipment. ABB ACS800-07, is available in sizes well above 1MW.

http://www.abb.com/product/seitp322/13f7f7a10dfd1e78c1256ddb002d0b1b.aspx?country=US

This stuff is not at all unusual on electric train systems & industrial drives. In your application you will simply be converting to fixed 60Hz & synching in.
Your DC battery bank will likely be organized to 800V or thereabouts. The drives integral controls will take care of damn near everything. A 1MW ACS800 might set you back $100k CAN or so, + engineering & installation costs.

Not an ad, but your regional ABB engineering office will be able to put you onto a local engineering outfit that can provid you with support through commisioning. Where in Canada are you?

Your local utility will take care of transformation, re-distribution & protection. 1MW is pretty small stuff for a utility, so it will only involve engineering as far as the nearest sub-station. You just need to provide a 600V, 1000A 3-ph service.

=======

Why go direct PV electric at all?
1MW would be much easier to obtain via solar-thermal. Heat-pump up to generate usable steam to turn a conventional rotary synchronous alternator.
The panel array would need to be 1250 m2, but you can get by with a whole lot less if you use tracking mirrors, and even avoid the messy heat-pump.

M

[Mike90250]

Why go direct PV electric at all?

maybe something to do with the difference of washing off the panels, vs other more complex methods.