That Decentralised Low Voltage Local DC Power Grid, How Did It Do?

Early on in the year, Hackaday published one of its short daily pieces about plans from the people behind for a low voltage DC power grid slated for the summer’s SHACamp 2017 hacker camp in the Netherlands. At the time when it was being written in the chill of a Northern Hemisphere January the event seemed so far away, but as the summer fades away along with the deep tan many SHACamp attendees gained in the Dutch sunlight it’s worth going back and revisiting the project. Did they manage it, and how did they do? This isn’t really part of our coverage of SHACamp itself, merely an incidental story that happens to have the hacker camp as its theatre. 

A Working DC Power System

The SHACamp 2017 solar array when the sun was out.

As someone with an interest in DC power their village was high on my list of priorities even if its distance from the UK hackspaces village was too high for me to have my own supply. What I found when I made my way over in the persistent rain of the first day was a fully functional DC power grid, albeit not as ambitious a deployment as that planned when first we reported on the project.

Power came from a set of solar panels, each with an onboard regulator to the grid’s 42 volts. These were concentrated in a central distribution cabinet, from which was fed a sub-cabinet for distribution to villages and other users. The plan for a full-scale grid is to have a network of such cabinets, however the huge logistical challenge that posed for a team of volunteers on their first deployment meant that the system on show was considerably more modest.

Peering inside the cabinets revealed arrays of circuit breakers and custom boards for current monitoring and voltage regulation. The cabling in use was not of the size you might expect, because of the relatively humble size of the installation it appeared that they were using hefty mains cable as the current in question was within its capability.

Regulator and charger on the other end of the wire in the ChaosWest village
Regulator and charger on the other end of the wire in the ChaosWest village

It was time to follow one of the cables snaking a few hundred metres across the grass of Scoutinglandgoed, which took me to the Chaos West village, and incidentally to an encounter with an awesome propane campfire and some of the best popcorn I have ever tasted, but that’s another story. There I found their endpoint, and a home-built regulator providing a step-down to a variety of voltages that they appeared to be using for both LED lighting and to charge the batteries for their array of electric vehicles. They had a reported voltage drop of over 10 volts over their cable run, an inevitable result of both its length and comparatively small cross-sectional area of copper. They did however have a useable power supply, and were taking advantage of it.

It was clear from the SHACamp 2017 DC power grid that this was a fully functional operation able to safely generate and distribute electricity to the attendees in a manner that they could take advantage of for simply the cost of a commercial switch-mode converter board. It was obvious though that this was a limited deployment within the logistics available to them in the run-up to this event, but what it gives us is a preview of what we can expect at future events.

A Viable Proof Of Concept

In particular the key word missing from the SHACamp deployment is in the title of this piece. Decentralised. On this occasion there was only what amounted to a power supply node, or a power station if you will, and customer nodes. You might say that this was the very model of a centralised power system, the shortcomings of which were laid bare in an obvious manner by the voltage drop at the Chaos West village. Happily though it’s important to stress that this is not a problem of design but one of the logistics of this first deployment. The system is designed from the ground up to be supplied from an array of power sources from “official” power nodes to users feeding their own power back into it, and this offers us the exciting prospect of future camps featuring a more extensive and robust DC power infrastructure building on this one but with less concern over voltage loss.

When reviewing the DC power grid at SHACamp 2017 it is therefore best not to criticise it for any shortcomings it may have had. Instead we should view this outing as a proof of concept of an exciting development that will feature in future camps. It is not a demonstration of the limits of what the people can do but a taster of what is to come, and that can only be exciting.

More information about the grid including significant technical detail can be found at

25 thoughts on “That Decentralised Low Voltage Local DC Power Grid, How Did It Do?

  1. The network diagram in the presentation they originally shared didn’t make any sense (e.g. why would you use redundant conductors if you aren’t going to make proper use of them?) but the project is very interesting nonetheless. The parallel connection of power sources is not straightforward unless they’ve been designed for that purpose. In general, DC supplies with ugly line regulation tend to behave nicely to each other when in parallel, while tight regulation leads to uneven load share and other issues.

    1. The redundant connections serve to some extent lower the path resistance by letting you “share” your neighbor’s line to the hub.

      But as for safety, such a mish-mash creates multiple possible feed paths to keep an electric arc going, so fuses and breakers don’t really work because the current may be coming from any which way.

      1. That’s the thing, it makes the protection harder with little gain. I’d rather put less wiring. Also in terms of losses, there should be way more power being lost in the dc/dc converters than in the wiring.

  2. Ahh… Alt-“Makers” making an Alt-Power thing – that doesn’t work. The next attempt will be “Makers” making useless Alt-Solar Roadways (once again). Sheesh. Then they say, “All of the specifications, hardware and software will be made available on this website in time.” Yeah right, “in time”, AFTER the Alt-Media is through with their Alt-Factless global warming Click-Bait posts. Once the “Green” hype dies down in the media coverage, there will be NO focus on the epic Alt-FAIL results that may come out “in-time”.

    1. Those electrical engineers missed their HVDC course, all the news about ABB’s dc datacentres, all the news about usb3 power delivery and what’s going on in electrification of the developing world.

  3. DC ‘micro’ grids, to be of any practical use, must operate at least 300V. As others mentioned, DC current fault interrupt is serious and dangerous stuff. And most new grid-connected PV strings typically provide 400V to 1500Vdc, which is why the scoped safety standards and code for this stuff (such as IEC62109-x, ANSI1741, ANSI1703, NEC article 690) have been re-defined and scoped for up to 1500Vdc.

    Power distribution stuff is not for the ill-informed. Mistakes that cause problems will allow your insurance company to abandon you and the local gendarmes will probably arrange for an indictment on a charge or two or three.

    Note that the Klingon empire has no relevant electrical safety standards or building code, as they tend to allow their weak and/or stupid to self-destruct.

    1. Depends on the area you’re trying to cover. 120 Volts is enough for modest distances.

      One thing that is not often appreciated is that DC power flows through plain iron bailing wire, where AC does not, due to the skin effect. Saves quite a bit on material costs.

      1. Quite frankly, for 100m distances the difference between ac and dc in the losses of the conductors is negligible and much smaller for example than the power lost in a cheap dc/dc converter.

        The benefit of dc at this scale is the ease of conversion using cheap dc/dc as opposed to inverters and hesvy transformers. Not a big deal to use transformers in a fixed location but these nomads in the article need lightweight solutions.

  4. “When reviewing the DC power grid at SHACamp 2017 it is therefore best not to criticise it for any shortcomings it may have had. ”

    Oh Jenny! That is like issuing a challenge to the commenters to criticize at Full Throttle!

  5. Correct me if I am wrong but when you lose 10 volts to the resistance of your distribution system, that like losing %20-25 of your available power and just heating up your service line. Isn’t a typical power distribution system designed to keep this in the low single digit range?

      1. Just tripling the line voltage reduces the losses to 11% of that 24% = 2.6%, as they’re I^2R losses.
        But 126 volts is no longer ELV, so can’t legally be tackled by muffins. And not only are DC arcs more difficult to extinguish (e.g. need HRC fuses), but DC is also more likely to stop your heart. It’s definitely “one hand in your pocket” time, and “never work on it when it’s live if you wan to too”.

        It is now almost half a century since I read of a HV DC power link between Sweden and Denmark. Similarly, a professionally manufactured bidirectional HV DC-DC converter at line portals between DIY low voltage local subnets is needed to make the concept practical even intra-village.

        Here in Australia, where we have nearly 2 km to the one neighbour in sight, it’s far more practical to remember what Nikola Tesla indicated long ago – there’s free energy in the ether – it’s falling on your head when outdoors during the day. So forget the wire. Just put up panels to catch the bounty.

    1. It was a very long cable run because there was only the one node. And to be honest the cable wasn’t thick enough.

      The voltage by the way is because of the European Low Voltage Directive. Under IIRC 60VDC or 50VAC you can do what you want, above that and all sorts of laws kick in.

  6. Oh, didn’t see this new article, thanks Jenny!

    Well, first of all, it was a test, some things went perfect, some went not that good, but that’s what testing is about.
    But in the end it’s a relatively cheap way to share energy over short distances safely. Continuing this! :)
    But it’s going to take time, it’s a hobby project, most of it is done by a single person. So updates will *eventually* come and not really frequent.

    Voltage drops were expected and calculated beforehand, no surprises there. And it improved a bit when people switched their 60m CCA cable with a copper one. But still, it’s not good for longer distances at high currents and it’s not designed for that either.

    The next step is building a few lithium battery packs for storage. This way, the *not so clean* 400V->42V PSU is not needed anymore. Well, until the batteries are empty of course. And what’s also on the list is a bi-directional DC-DC converter, 10-60V on both sides for connecting your “local” 12/24/whatever system to it.

    Oh and here’s an article one of the participants wrote:

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.