Ultimately, the problem with solar, wind, and any other "periodic" or localized generation method is storage and distribution. I live in an area with incredibly cheap power due to hydro dams that generate well in excess of what we need. However, there's only so much we can do with that excess in terms of efficiency of generation because it costs so much to send that energy long distances.
In addition, you have the obvious periodic nature of solar and wind, which mean that not only do you have the distribution problems, but you also can't just flick it on and offset a bunch of peak load or burst demand (such as when hospitals "turn on" in the morning).
You could have 100% efficient solar technology, and it wouldn't change much the economics of green energy because of the physical nature of problem. Thus, you need to change how we deal with the physical nature, and thus the economics themselves by improving the technology in other places. I work for a startup which is building a combination of efficient storage infrastructure and a distributed software system to help manage the heuristics of controlling supply and demand with this storage system and generation technology (including solar and wind, but also traditional generation as well), helping to do things like smart "time shifting" of energy supply.
Since superconductors are still a ways off, and even with them it doesn't change the nature of periodic generation, I think storage technology is likely the most important area for investment in the immediate future (and not just because that's where my paycheck comes from :) )
That sounds like a great start up. Are you allowed to clarify what you mean by efficient infrastructure? That is, do you mean the topology of the network, the routing on the links or the type of nodes? What type of algorithms do you use for load balancing? Are you storing on more efficient batteries for example? What is the chemistry of your batteries? Or are they the dual, fuel based?
Sorry, this has been a bit of a ramble just curious about your system.
He stresses that financing, generation and distribution must all be improved in sync for a successful transition to occur. This is based on his experience in Europe helping them transition to distributed renewable energy.
It's funny you mention that - that book has been making its way around at work, and I've got it sitting on my desk to read right now.
I think the author is right on, though I bet we'll see some significant incremental progress in the next couple years. I also think, based on the blurbs I've seen, that he's missing out on one key part of the equation that a lot of non-energy startups can go after - intelligence in the system. We'll be building the distribution, storage and generation side, and the smart grid will continue to plug into more and more devices, but it's the smarts of the system at scale, doing things like we've all been working on in the social space for a few years now, that will take those raw materials to the next level, and do something really clever that none of us have thought about yet.
For me, the biggest surprise in the interview was the reaction of the legacy energy companies. At first, they were very unhappy with the idea of independent, distributed generation (no surprise). But once it became clear that A) It's happening with or without them. And B) There's a new role for them in distribution intelligence. They got on board and now are actually pushing progress forward.
It's a great company to work for. Lots of smart, motivated folks with tons of experience both in the energy industry and other connected industries. I'm in the second wave of new hires, working as an architect on the software system that will be managing the scale of the infrastructure as we build out. That said, my knowledge is obviously fairly specific to the software, and the hardware bit that I've come to understand is primarily through osmosis :) To try and answer what I can without violating the usual secret squirrel stuff:
Efficient infrastructure - we're able to charge and discharge at a high level of input to output efficiency. Typically, in the energy storage business, you lose a lot of energy via inverters and other technology. Our secret sauce gets around a lot of this, though I don't think I'm able to disclose exactly how it works. It's incredibly elegant in it's simplicity though, and can plug directly into the wild west of the grid. We're also able to network together a significant amount of capacity in almost any configuration for a very reasonable cost per KW. That means we can take our storage system and sell you 80kw/h discharged over 4 hours (20kw per hour), or 40kw over 2 hours, or just keep adding capacity to change those numbers further. While we're a bit young to be just "plug and play", we'll get there as time goes by.
In terms of battery technology, we have a few tricks up our sleeves and are working with a variety of researchers, but for the most part we just work with what's on the market - meaning high end lead acid (best bang for the buck, and the maintenance story is improving), and lithium ion where people want it. This helps keep purchase and maintenance costs low, but we've also built our own battery management system that lets us plug and play any battery tech on the market today, as well as several that are on the way but haven't hit a decent economy of scale yet.
I can't comment on specifics of the BMS, other than we're building the hardware as well as the software, which lets us play with a variety of models for handling battery input and output.
It's a market that's seeing plenty of entries and new competition every day, but it's also really cool to see how people are approaching the problem and working with partners to drive the cost of energy down. That can only be a good thing for all of us!
On the other hand, and from what I remember from a conversation with someone who works on the UK National Grid, Solar and Wind power have the advantage of predictability; which in many ways is just as important as availability for large-scale power infrastructures. (That's short-term predictability, generally < 1 hour).
As they've been saying since the 1950s, fusion is only 20 years away to solve all our problems anyway.
Oh, I don't mean to imply that solar and wind are entirely unpredictable (I was hoping my comments about hydro would cover that), but they aren't always predictable when and where you need it. One of the problems for BIG infrastructure (say, hospitals and apartment buildings), is that they often have their peak demand almost instantaneously in the morning - often before the sun comes up. For solar, that means you can't directly address this peak demand without some way of shifting that generation capacity. But, as you say, unlike some of the arguments over the years trying to prove that solar is untenable, I think it's predictability, combined with additional advances elsewhere, will indeed push us toward a better future, at least until that "fusion" stuff gets around to being feasible :)
The electricity distributer in South Australia is able to control the compressor pump of a wide area of the population's air conditioners. They use this during high demand periods (like hot days in summer) to reduce the loading on power lines.
People don't really notice that their compressor pump is being cycled to reduce system demand. Are you working in this area too?
> During high summer energy use periods, PECO Smart A/C Saver works to reduce electricity usage by "cycling" central air conditioners throughout the region
> Using around 1000 volunteer households, trials to date indicate a 19 - 35% reduction in peak load where direct load control demand management is utilised.
There is a 19 - 35% percent reduction in power usage using the technique. Payment for opting into the program looks like a great idea. Could the concept be extended to freezers and fridges? Imagine if you controlled a network of these small appliances and sold the ability to reduce peak demand to your distribution company.
I can't speak to overall strategy, as that's above my pay grade, but I can say that I HAVE worked in that area and I think everybody here is really interested in getting into that.
At my last startup, we had a prototype going at my house where we had a specialized smart meter with relays attached to the hot water heater, and an internet enabled thermostat. We could monitor, control and schedule peak load over the course of the day pretty easily. It was pretty labor intensive to set everything up (it was mainly the tapping of the mains and running the relays to the hot water heater), but actually worked well based on some simple schedules. I estimated that I saved about 15 to 20% of my power bill once it was in place.
The great thing about all of this, is that like most internet applications, the benefits become more pronounced and easier to achieve at scale. The more you can feed into a system, from both the supply side and the demand side, the better you can get at predicting and controlling the nature of the grid. And that transitions into some pretty serious effects on the economics of the situation, without really needing any major technology breakthroughs.
Comments
Ultimately, the problem with solar, wind, and any other "periodic" or localized generation method is storage and distribution. I live in an area with incredibly cheap power due to hydro dams that generate well in excess of what we need. However, there's only so much we can do with that excess in terms of efficiency of generation because it costs so much to send that energy long distances.
In addition, you have the obvious periodic nature of solar and wind, which mean that not only do you have the distribution problems, but you also can't just flick it on and offset a bunch of peak load or burst demand (such as when hospitals "turn on" in the morning).
You could have 100% efficient solar technology, and it wouldn't change much the economics of green energy because of the physical nature of problem. Thus, you need to change how we deal with the physical nature, and thus the economics themselves by improving the technology in other places. I work for a startup which is building a combination of efficient storage infrastructure and a distributed software system to help manage the heuristics of controlling supply and demand with this storage system and generation technology (including solar and wind, but also traditional generation as well), helping to do things like smart "time shifting" of energy supply.
Since superconductors are still a ways off, and even with them it doesn't change the nature of periodic generation, I think storage technology is likely the most important area for investment in the immediate future (and not just because that's where my paycheck comes from :) )
That sounds like a great start up. Are you allowed to clarify what you mean by efficient infrastructure? That is, do you mean the topology of the network, the routing on the links or the type of nodes? What type of algorithms do you use for load balancing? Are you storing on more efficient batteries for example? What is the chemistry of your batteries? Or are they the dual, fuel based?
Sorry, this has been a bit of a ramble just curious about your system.
I really appreciated this interview http://itc.conversationsnetwork.org/shows/detail5059.html with the guy who wrote the book http://en.wikipedia.org/wiki/The_Third_Industrial_Revolution
He stresses that financing, generation and distribution must all be improved in sync for a successful transition to occur. This is based on his experience in Europe helping them transition to distributed renewable energy.
It's funny you mention that - that book has been making its way around at work, and I've got it sitting on my desk to read right now.
I think the author is right on, though I bet we'll see some significant incremental progress in the next couple years. I also think, based on the blurbs I've seen, that he's missing out on one key part of the equation that a lot of non-energy startups can go after - intelligence in the system. We'll be building the distribution, storage and generation side, and the smart grid will continue to plug into more and more devices, but it's the smarts of the system at scale, doing things like we've all been working on in the social space for a few years now, that will take those raw materials to the next level, and do something really clever that none of us have thought about yet.
For me, the biggest surprise in the interview was the reaction of the legacy energy companies. At first, they were very unhappy with the idea of independent, distributed generation (no surprise). But once it became clear that A) It's happening with or without them. And B) There's a new role for them in distribution intelligence. They got on board and now are actually pushing progress forward.
It's a great company to work for. Lots of smart, motivated folks with tons of experience both in the energy industry and other connected industries. I'm in the second wave of new hires, working as an architect on the software system that will be managing the scale of the infrastructure as we build out. That said, my knowledge is obviously fairly specific to the software, and the hardware bit that I've come to understand is primarily through osmosis :) To try and answer what I can without violating the usual secret squirrel stuff:
Efficient infrastructure - we're able to charge and discharge at a high level of input to output efficiency. Typically, in the energy storage business, you lose a lot of energy via inverters and other technology. Our secret sauce gets around a lot of this, though I don't think I'm able to disclose exactly how it works. It's incredibly elegant in it's simplicity though, and can plug directly into the wild west of the grid. We're also able to network together a significant amount of capacity in almost any configuration for a very reasonable cost per KW. That means we can take our storage system and sell you 80kw/h discharged over 4 hours (20kw per hour), or 40kw over 2 hours, or just keep adding capacity to change those numbers further. While we're a bit young to be just "plug and play", we'll get there as time goes by.
In terms of battery technology, we have a few tricks up our sleeves and are working with a variety of researchers, but for the most part we just work with what's on the market - meaning high end lead acid (best bang for the buck, and the maintenance story is improving), and lithium ion where people want it. This helps keep purchase and maintenance costs low, but we've also built our own battery management system that lets us plug and play any battery tech on the market today, as well as several that are on the way but haven't hit a decent economy of scale yet.
I can't comment on specifics of the BMS, other than we're building the hardware as well as the software, which lets us play with a variety of models for handling battery input and output.
It's a market that's seeing plenty of entries and new competition every day, but it's also really cool to see how people are approaching the problem and working with partners to drive the cost of energy down. That can only be a good thing for all of us!
On the other hand, and from what I remember from a conversation with someone who works on the UK National Grid, Solar and Wind power have the advantage of predictability; which in many ways is just as important as availability for large-scale power infrastructures. (That's short-term predictability, generally < 1 hour).
As they've been saying since the 1950s, fusion is only 20 years away to solve all our problems anyway.
Oh, I don't mean to imply that solar and wind are entirely unpredictable (I was hoping my comments about hydro would cover that), but they aren't always predictable when and where you need it. One of the problems for BIG infrastructure (say, hospitals and apartment buildings), is that they often have their peak demand almost instantaneously in the morning - often before the sun comes up. For solar, that means you can't directly address this peak demand without some way of shifting that generation capacity. But, as you say, unlike some of the arguments over the years trying to prove that solar is untenable, I think it's predictability, combined with additional advances elsewhere, will indeed push us toward a better future, at least until that "fusion" stuff gets around to being feasible :)
The electricity distributer in South Australia is able to control the compressor pump of a wide area of the population's air conditioners. They use this during high demand periods (like hot days in summer) to reduce the loading on power lines.
People don't really notice that their compressor pump is being cycled to reduce system demand. Are you working in this area too?
I'm in the US and my electric company has that too. They pay up to $30 per month in the summer if you opt-in to the program.
https://www.peco.com/Savings/ProgramsandRebates/Residential/...
> During high summer energy use periods, PECO Smart A/C Saver works to reduce electricity usage by "cycling" central air conditioners throughout the region
The trial [1] I saw suggested that:
> Using around 1000 volunteer households, trials to date indicate a 19 - 35% reduction in peak load where direct load control demand management is utilised.
There is a 19 - 35% percent reduction in power usage using the technique. Payment for opting into the program looks like a great idea. Could the concept be extended to freezers and fridges? Imagine if you controlled a network of these small appliances and sold the ability to reduce peak demand to your distribution company.
[1] http://www.etsautilities.com.au/centric/our_network/demand_m...
I can't speak to overall strategy, as that's above my pay grade, but I can say that I HAVE worked in that area and I think everybody here is really interested in getting into that.
At my last startup, we had a prototype going at my house where we had a specialized smart meter with relays attached to the hot water heater, and an internet enabled thermostat. We could monitor, control and schedule peak load over the course of the day pretty easily. It was pretty labor intensive to set everything up (it was mainly the tapping of the mains and running the relays to the hot water heater), but actually worked well based on some simple schedules. I estimated that I saved about 15 to 20% of my power bill once it was in place.
The great thing about all of this, is that like most internet applications, the benefits become more pronounced and easier to achieve at scale. The more you can feed into a system, from both the supply side and the demand side, the better you can get at predicting and controlling the nature of the grid. And that transitions into some pretty serious effects on the economics of the situation, without really needing any major technology breakthroughs.