Path: ...!weretis.net!feeder9.news.weretis.net!news.quux.org!eternal-september.org!feeder2.eternal-september.org!news.eternal-september.org!.POSTED!not-for-mail
From: john larkin <JL@gct.com>
Newsgroups: sci.electronics.design
Subject: Re: 1GW (sic) Battery Energy Storage Systems
Date: Mon, 25 Nov 2024 08:49:44 -0800
Organization: A noiseless patient Spider
Lines: 81
Message-ID: <sca9kjd1nttphgdftg8o76qlt920k0l708@4ax.com>
References: <vhv3b0$27b8q$2@dont-email.me> <pfn6kjl8ti6kuf31nb8ot7io5pvpkdurhr@4ax.com> <ql69kjhe7ikb5h76v9mcagcebcj7podqnr@4ax.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Injection-Date: Mon, 25 Nov 2024 17:49:48 +0100 (CET)
Injection-Info: dont-email.me; posting-host="bb2d0afb0008a25ee556e1c1ca9aa451";
	logging-data="3024112"; mail-complaints-to="abuse@eternal-september.org";	posting-account="U2FsdGVkX18dsjU/vLWM02l2O52QZkx4"
User-Agent: ForteAgent/8.00.32.1272
Cancel-Lock: sha1:4yiTHvy6tZ700wTttVqW+u5Biew=
Bytes: 4483

On Mon, 25 Nov 2024 11:20:07 -0500, Joe Gwinn <joegwinn@comcast.net>
wrote:

>On Sun, 24 Nov 2024 09:21:13 -0800, john larkin <JL@gct.com> wrote:
>
>>On Sun, 24 Nov 2024 11:40:16 +0000, Martin Brown
>><'''newspam'''@nonad.co.uk> wrote:
>>
>>>There is a startup company in the UK called NatPower (no relation to the 
>>>real National Power PLC aka NPower) claiming to have over 25 years of 
>>>experience in BESS. They are proposing 1GW (sic) BESS in all sorts of 
>>>brown and green field UK sites. They have no track record that I can 
>>>find. Does anyone here know of any real international project(s) that 
>>>they have actually done and more importantly *delivered*?
>>>(their own PR or its uncritical repetition in the press doesn't count)
>>>
>>>Their website is very slick indeed but I suspect that beauty is only 
>>>skin deep. I am interesting in evidence of substance not PR fluff.
>>>
>>>Can someone provide me with some rough estimates of what a 1GW class 
>>>energy storage system ought to look like in terms of layout and the 
>>>number of container sized units and space required?
>>>(or at least sanity check my guesstimate below)
>>>
>>>Assuming that they mean 1GWh I reckon it will be about 2000T of Lithium 
>>>batteries. If each module is storage container sized and can contain 
>>>20m^2 of batteries I reckon it is about 100 units (twice that if I have 
>>>over estimated how much battery you can safely fit in a module). I 
>>>haven't been able to find any manufacturers specifications for them.
>>>
>>>At what point in the scale up from 50MW storage units (which are quite 
>>>common in the UK) to these new Gigaparks do things get interesting?
>>>
>>>It also strikes me that if these storage battery systems have similar 
>>>characteristics to the Lithium ion cells in my laptop they will require 
>>>complete replacement every 5 or so years if they get cycled daily.
>>>
>>>And I presume each module needs integrated fire suppression systems to 
>>>handle thermal runaway problems. Lithium fires being notoriously 
>>>difficult for ordinary fire fighting methods to put out.
>>>
>>>Also what additional measures will it need to tie into 400kV supergrid?
>>>
>>>How do you even do that at 1GW using semiconductor components?
>>>
>>>It can obviously be done since some big interconnectors are DC but how 
>>>much does that sort of hefty high voltage infrastructure cost?
>>
>>Maybe they mean 1 gigawatt. It seems to be peaking power.
>>
>>There are already long-distance DC power lines, with AC conversion on
>>both ends. I think they use a lot of IGBTs.
>
>Yes.  The key is optical triggering via an optical fiber, which is
>both precise in time and electrically isolated, so one can stack many
>units in series.  
>
>The highest-power stuff is Thyristor based, and is directly optically
>triggered.
>
>IGBTs are triggered by an optically-triggered IGBT driver.
>
>
>>Imagine firing up a 12 gigawatt  800 KV converter for the first time!
>
>Done outdoors in staged steps, from a safe distance.
>
>There will be many wires.  12e9/8e5= 1.5e4 amps total.  Icing will not
>be a problem.
>
>
>Joe Gwinn

IGBTs are interesting but slow!

We use SiC and GaN for fast high voltage stuff, but to us high voltage
is hundreds of volts or low KV.

People are making GaN fets with integrated drivers. Optical input
would be cool.