Author Topic: Looking for EE Partner(s) to start an open source hardware project / company  (Read 3612 times)

0 Members and 1 Guest are viewing this topic.

Offline snickTopic starter

  • Contributor
  • Posts: 13
  • Country: gb
Replaying on where the 30% comes from:

- the 30% is in order to control the 3 big loads of the home of the future (EV charging (both mono-directional than v2h / Battery Charing / Heat pump for heating). Note for the non Europeans (EU has set aggressive goals for transitioning to heatpumps, with an expected inventory hitting the market of ~200M units in the next 4 years).
All the rest is not to be controlled (you are correct, nobody wants to time their washing machine.

- What is surprising (and that's where ML comes in) is that a lot of people cook always at the same time, and wash things often at the same time.

- it is true that energy companies are looking at a more dynamic pricing. I worked with octopus energy in UK to prove that a heatpump can cost to the user 70% less money with the same heating comfort, if a 3 range tariff was to be implemented. With no major downside for the energy company (because of wholesale prices).


On SwitchDIN. I think that is indeed a good part of what I had in mind. It's great for monitoring, I'm not sure how much they can run the controlling part (can they tell an inverter where to send energy? can they turn on / off a heatpump? ). I'm impressed they have the tech and implementation to already do VPPs.
I'd be VERY interested to understand they 'go to market' strategy. How do they get to more homes.





 

Offline f4eru

  • Super Contributor
  • ***
  • Posts: 1086
  • Country: 00
    • Chargehanger
Quote
(you can save up to 30% energy)
Nope you cannot.
Time shifting loads does not save any energy.
Also, putting active boxes everywhere that draw a few watts each constantly does actively worsen the power bill.
 
The following users thanked this post: Someone

Offline Phoenix

  • Frequent Contributor
  • **
  • Posts: 420
  • Country: au
On SwitchDIN. I think that is indeed a good part of what I had in mind. It's great for monitoring, I'm not sure how much they can run the controlling part (can they tell an inverter where to send energy? can they turn on / off a heatpump? ). I'm impressed they have the tech and implementation to already do VPPs.
I'd be VERY interested to understand they 'go to market' strategy. How do they get to more homes.

I don't know which manufacturers they have integration with for loads. But as long as there is an interface they are willing to work with all sorts of manufacturers for integration. Maybe load interfacing is part of their future plans - the control interfaces are usually pretty trivial to implement.

Quote
can they tell an inverter where to send energy?

You can't tell an inverter where to send energy, as such.
- You can tell an inverter to generate more or less power. This is how their VPP's work - controlling the stationary battery inverter.
- You can tell loads to consume more or less power.

Quote
How do they get to more homes.

They really don't work with consumers directly. They work with the utilities, the OEMs, installers etc. Their business model is around access to the cloud data and control interface. You can get paid by the consumer to access their data/interface and get paid by the utility to access the same data :D.
 

Offline nctnico

  • Super Contributor
  • ***
  • Posts: 26755
  • Country: nl
    • NCT Developments
I agree. I don't really see any use for consumers to have fine grained control over loads. This is more something that is interesting to grid operators but even then this will somehow impact consumers. Over here grid companies used to (or still do) send signals over the grid when the cheap night tariff starts and ends to switch on electric water heaters. This makes it convenient for the consumers so they don't have to remember. But these electric water heaters have become very rare because in the end having a limited hot water supply is just cumbersome to work with. IMHO a solution that aims to even out loading of the grid better should also benefit the users in terms of comfort otherwise it is just a step back. And that brings me to why I think that in the long run these systems have no future.

It looks like the grid companies are going to implement large scale storage one way or another. Why would they sell electricity cheap if they can sell it for a higher price later on? Grid scale storage will always be cheaper compared to home storage. So long term prices will even out. Before Putin got crazy and invaded Ukrain, I paid the same price for the day and night tariff. This means that the energy companies didn't need to dump excess electricity.
There are small lies, big lies and then there is what is on the screen of your oscilloscope.
 

Offline snickTopic starter

  • Contributor
  • Posts: 13
  • Country: gb


Quote
You can't tell an inverter where to send energy, as such.


On SwitchDin I don't see any support for BMS, so how do they handle the following case:
Solar inverter generating 1kWh, battery is at 30% and can take 7kWh to top up, grid tariff is best price for sell energy to. 
Optimal outcome, don't charge the batter, just sell energy to the grid. 
 
 

Offline snickTopic starter

  • Contributor
  • Posts: 13
  • Country: gb
Quote
(you can save up to 30% energy)
Nope you cannot.
Time shifting loads does not save any energy.
Also, putting active boxes everywhere that draw a few watts each constantly does actively worsen the power bill.

Ok, I give you that '30% energy' here really means (30% in cost, either by consuming less energy, or being smarter in when you purchase it).
But on consuming less energy: 

- time shifting is not the only control optimisation you can run.
Peak trimming works extremely well for heat pumps, especially once you have a good understanding of your close control loop (aka you know how long it takes to reach a temperature X in a determined environment). The COP (power consumption over energy generated) of a HP is non linear, and if you have high control of where you are you can substantially change consumption at equal comfort. 



 

Offline nctnico

  • Super Contributor
  • ***
  • Posts: 26755
  • Country: nl
    • NCT Developments
I think you'll find that the housing market is about to collapse globally. I would be weary of putting in any investments in technologies that require you to first own a home.
Why would the housing market collapse? Tons of people are looking for a home. Besides that, companies renting homes can also benefit from improving their homes to be less costly to live in.
There are small lies, big lies and then there is what is on the screen of your oscilloscope.
 

Offline Phoenix

  • Frequent Contributor
  • **
  • Posts: 420
  • Country: au
On SwitchDin I don't see any support for BMS, so how do they handle the following case:
Solar inverter generating 1kWh, battery is at 30% and can take 7kWh to top up, grid tariff is best price for sell energy to.
Optimal outcome, don't charge the batter, just sell energy to the grid.

1. The Droplet has 2 way control established with the solar inverter and ESS.
2. The cloud system takes in all the parameters it can (solar generation, current house usage, battery capacity, time-of-use tariff, expected load profile, weather forecast etc.).
3. The cloud sends the "charge" or "discharge" power command to the ESS via the Droplet.

(Note that their list of integrations include both solar inverter and ESS).

How would your proposed system perform this?
 

Offline snickTopic starter

  • Contributor
  • Posts: 13
  • Country: gb
1. The Droplet has 2 way control established with the solar inverter and ESS.
2. The cloud system takes in all the parameters it can (solar generation, current house usage, battery capacity, time-of-use tariff, expected load profile, weather forecast etc.).
3. The cloud sends the "charge" or "discharge" power command to the ESS via the Droplet.

(Note that their list of integrations include both solar inverter and ESS).

How would your proposed system perform this?

You are correct. It's load side driven control. My proposed system would work exactly in the same way, with no data leaving the premise, just data coming in. 
I had a very hard time seeing what ESS droplet was supporting from the website (beyond the battery icon next to their architecture diagram).

Look, kudos to them, they have done 90% of what I want to do. I hope they are getting the traction they deserve. I'll try to reach out to them and hear if there is any opportunity to collaborate. 


 

Offline Bicurico

  • Super Contributor
  • ***
  • Posts: 1707
  • Country: pt
    • VMA's Satellite Blog
I don't have any PV equipment installed. A collegue of mine has and his ROI was about 10-20 years (the goverment changed the incentive meanwhile, making it much less profitable).

And I am not that literate in how to implement a home PV equipment.

But in the german magazine C'T there have been recently some articles dedicated to PV solutions.

The main points are:

1) You may not just connect a PV system to the grid. For that to happen, you need to have a professional installation, which certainly needs to use certified equipment.
2) You may use a private PV system not connected to the grid, but only up to a given dimension. In this case you have the problem that you need to put the generated power in use, store it to batteries or dissipate it as heat.
3) Batteries are expensive.
4) Doing the math, I really doubt that you get any money out of a PV system where you can get regular power from the grid. The PV systems are still expensive: the panels might have dropped considerably in price, but the inverter, power management system, battery, supports, etc. still sum up considerably.
5) We are currently facing inflation: paying the credit of a house is getting considerably more expensive, like all other goods. Personally, I started to spend less money and making sure I will be able to face 2023-2025 without financial troubles. But definitely I don't have spare money for a PV system.
6) The ROI of a PV system is worse than the ROI of financial products (shares, gold, whatever): while you can sell those (even at some loss) in case you need the cash, you won't be able to sell the PV system. I get it that Germans in particular (I am German, too), have this doctrine of wanting to save mother earth and single handed want to face the CO2 emission. But that is definitly not the goal of most other EU citizens - they just want to have cheap energy).
7) If I was to invest in a PV solution, I would just order the best commercially available offer, which in Portugal is offered by the main electric company themselfes - with credit and all! I would not care the least about open source in hardware or software. I would care about price, quality, ROI and technical support, including spare parts.

So with all what I said, I would REALLY like to understand what this project is supposed to bring? I am not mockinig nor do I want to sound offensive at all. I would simply like to understand, what I am overseeing!

It does not make sense at all. Where is the need for any hardware standard? The panels produce power, it does to the management module (many different ones in the marked) and then to an inverter and/or battery. What is there to be compatible or incompatible?

Thanks for the patience in hopefully enlighten me.

And sorry for insisting that all of this doesn't make sense to me - I am really open to learn what I don't know.

Kind regards,
Vitor

Offline snickTopic starter

  • Contributor
  • Posts: 13
  • Country: gb
No worries Bicurico, let me try to explain the rational for this projects a little bit more in depth. Sorry for the stream of words, it would be easier to do a live session. 

You can think this project as composed of 2 parts. The optimization part and the system part.

Let's start from the optimisation part.

The current set of systems out there are pretty dumb and not interoperable. This means usually two things 1) you get locked in with a provider that overcharge you for that (ie.: Viessman who's selling PV at 3x because they can assure you they will work with their heatpump), or 2) you are ending up wasting more energy (or paying more for it). 
What do I want to have instead? each component of your home (at least the big loads - heat pumps, batteries, ev chargers) can communicate between them, can learn they operation patterns (when do you charge, what temperature do you want when, and can be aware of their context (ie.: if today is very sunny). 

Additionally, I'm betting on the fact that big retail energy operators are moving towards a dynamic pricing to better model the cost of energy at wholesale (right now in UK the wholesale price can 10x during the day). Assuming that is correct, you will have the cost of energy from the grid to change during the day. The same 1kWh will cost you anything from 0.05 euro to 1 euro (ballparking here). 

If this intuition is correct, your ability to decide when it's a good idea to buy from the grid, or to consume from your self generated energy will be critical. Equally, when it's worth to charge your battery/evs and when not.

The best effort (consume as much locally when possible) is suboptimal even in basic existing scenario like the two tarriffs system we have in UK. 
(you can read more here: https://octoenergy-production-media.s3.amazonaws.com/documents/MADE_High_Level_Design_PassivSystems.pdf)
  
This is basic demand-side response done locally, with a system that can produce and store energy. My calculation (and few other papers out there) shows that you can cut cost (in the current pricing market) by ~30% on a 200$ monthly bill. 

What are the practical problems to do this: 1) there is little interconnectivity between systems. ie.: the battery rarely knows how much the Inverter is generating from PV. 2) even if they were all connected, the current optimization mechanisms are very rudimental. We are looking at very similar to best effort optimisation mechanisms, they don't have awareness of pricing, they don't have awareness of environment condition like occupancy, irradiance, and temperature, they don't have any understanding of the user behaviour. Sure you can probably hack something together and have your HomeAssistant do something more advance, but your average user doesn't know what Home Assistant is. 


Ideally my idea stops here. I could go to the big player out there and say 'hey, we can build a better world, what about you give me access to your protocols, maybe you implement 2-3 more interfaces to get additional information from your load / system component, and in exchange you can be part of a bigger effort to decarbonize the residential space' . 
In an ideal world, everyone would say 'sure' , and what would be left to do is to build a tiny hardware that connects everything, and the software on top. All of a suddent all the home, that have already PV would be 'smarter'. 
Unfortunately, it doesn't work in that way, because, as said before, there are interest in keeping systems closed, even if this offers a suboptimal experience to the user. 


This is were part 2 kicks in. 
If incentives are not alined, force them. 

If hardware is not interoperable, build it. Leverage the consumer demand to drive the offer. 
If I need to go into the business of building inverters, I'll go into the business of building inverters, until someone understand that they can copy my design, manufacturer cheaper, and start flodding the market with more hardware that is compliant with my first part of the idea. 

How do I get the consumer to want my type of hardware ? simply by saving them money. If I can give users a 60$/month margin on their bill that's 7200 in 10 years, at parity of hardware cost (most likely mine will be more expensive, but still within a price range that makes this convenient for the user).
And for the users that don't have a system like that (no PV, no battery, etc), the value proposition needs to be that their total energy bill (ideally heating + electricity) needs to be higher than the cost of the system, over a 10-15 years horizon. A system like the one that I envision is probably a 18.000$ (7kWh generation, 10kWh storage, heatpump, installation) , including a heat pump, which means, if you spend more than 140$ a month in electricity and heating, you are better of with my system. 

You are right that the end consumer will hardly be aware of this. And having the right channel (ie.: your utility company being the reseller) is critical to reach scale. 
My plan is to get 10.000 residential units buying into this, at which point the entire system becomes more appealing for a retail energy provider (as they can do proper demand side response)
How I get to the first 10.000 residential unit is the tricky part, but that is marketing / sales / and connection (a small redevelopment in suburban london, has usually 7000 units, so you just need 1 partnership). What is really hard is the fact that I'm not an EE. and without a partner EE on this, there is no hardware, and if there is no hardware there is no software, and if there is no software there is no prototype to show to prove all my points 


 

Offline jonpaul

  • Super Contributor
  • ***
  • Posts: 3319
  • Country: fr
OP, your experience in running a business? tax, legal, regulatory compliance?

fund's for startup and unforeseen issues?

Sélection of a " partner" via a forum is unlikely to end well.


Finally 95% of all start-up business fail in first 5 years.

Just the ramblings of an old retired EE

Jon
Jean-Paul  the Internet Dinosaur
 

Offline snickTopic starter

  • Contributor
  • Posts: 13
  • Country: gb
Had 1 business before with a successful exit (sold it). Had few project started and killed within the first 2-3 months after reaching clarity on the details. did this both solo and with partners. 

I'm, in general, confident with running a business. I'm not confident in running a project like this by myself. Tax and legal are almost always outsourced. EE regulatory compliance is unknown to me.

My day to day work includes working on acquisitions and advising startups in the scaleup phase. I have access to capital and investor.Enough to give a ~10M runway to this before start making money (which I'm aware is not a lot).

I agree completely with you. I came here after running out of ideas. Coming from the software world there is little to no overlap with the EE community. I also feel the way people get to know each other is very different in the EE world compared to the software world.
I've also tried to tap into the local research center in EE at the main London engineering university, but again, they are used to work on big contract, and on pretty much advisory role.
Finally, I've been quite successful at getting in front of C-level in big electronic companies, but the lifecycle of their planning is incredibly long (5-years) and usually doesn't survive the leadership group. (For example I had an ongoing project with Schneider electric on this topic, which got canned after the CIO left). That's why I'm leaning towards a open source - ground up initiative.


ps.: interesting enough, Retired EE engineers would be the best fit for this project: experience, low concern about the risk, and (i guess) a lot of time to spare :D


« Last Edit: December 07, 2022, 06:52:12 pm by snick »
 

Offline nctnico

  • Super Contributor
  • ***
  • Posts: 26755
  • Country: nl
    • NCT Developments
Sélection of a " partner" via a forum is unlikely to end well.
Why? It is like dating. You won't find anyone unless you get 'out there'.
There are small lies, big lies and then there is what is on the screen of your oscilloscope.
 

Offline bidrohini

  • Regular Contributor
  • *
  • Posts: 201
  • Country: bd
You can post your requirement to www.freelancer.com
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf