Trimming wasted power with software defined electricity

[begreen]

Sounds promising. Has anyone been following this technology?
https://www.vox.com/energy-and-envi...ity-technology-efficiency-software-waste-3dfs

 

[DBoon]

In general, there are a lot of red flags here:

1. Someone who doesn't understand anything about the technology writing about technology and being very excited about how it is going to change the world.
2. Extensive use of metaphors (splashing water, pictures of turbulent water, etc.) probably given to them by some marketing guy at the technology company.
3. References to the lots of very smart people at the technology company who know what they are doing - they are not amateurs, of course, they've been around for a very long time and come from some reputable places, so you know they are good and very smart.
4. Lots of misstated facts about the underlying basis for why the technology is needed (admittedly, they correct the worst of these misstatements at the end of the article).
5. Lots of new made-up words and phrases (e.g. "Task-oriented Optimized Computing", or TOOC, "Flash Energy Storage System" or FESS, etc.)
6. Lots of buzzwords that sound impressive to non-technical people (e.g., 24-bit) and which the author has no idea the true meaning or relevance of.
7. Two endorsements by people who know what they are doing, because they sound excited and they have some knowledge in this area.

I could go on.

There is some merit to what they are doing, but the endless marketing hype and "we will fix the broken thing that's been perpetuated for 100+ years" and "this will save the world from climate disaster" hyperbole is a bit overblown. What they are doing is not too different from what a lot of other companies are doing with power electronics for grid applications, and the need for this equipment is a lot less than what they claim. If you want to improve transmission and distribution efficiency with power electronics, there are companies trialing products to utilities today that do this. This company seems to be aiming at the commercial market instead.

First, all the technology needed to do what they are doing has been around a long time - 24-bit digitizers, calculations of 24 different power parameters, real-time software calculations in FPGAs on streaming data, real-time power electronics systems to correct for power issues, etc. They've just simply packaged it up and are marketing it. Good for them - lots of good things come from people who package existing technologies into new applications to provide real value.

But what is their application? It seems that what will benefit from their power cleanup are loads at a customer site that rely on direct delivery and use of utility power without any intervening power conversion. Utilities deliver power with <5% distortion, so I doubt that the typical utility power is as bad as their graphs indicate, nor will "correcting" power at the customer site solve any of the transmission and distribution inefficiencies (roughly 7% of power is lost from transmission and distribution to delivery at customer site). And most electronic loads (and many motor loads, which represent half of power consumption) are supplied through some type of power conversion system (which takes the utility AC, creates DC, then inverts it back to AC at a specified frequency and voltage, or to another DC voltage, e.g., the server example). I don't see how any of those systems get any real benefit from this.

So, that leaves this system very well-suited to microgrid applications (perhaps - the ship example is one that I could see being very relevant) or very large industrial loads (which this doesn't seem targeted at).

There are others who post on this site who know more on these topics, so I look forward to their insights and comments.

 

[pdf27]

I work in aircraft power generation, and the problems this addresses are very relevant to us. However, there are other established ways of dealing with them that work really well - for instance, harmonics (which produce the distorted wave form - run a Fourier transform on the distorted waveform and you'll find it is actually a bunch of harmonics at 1x, 2x, 3x, etc. the grid frequency) can be very effectively suppressed by making small changes to the generator design at minimal cost and weight impact.
The really big potential market out there is if they can figure out how to cheaply insert it into say data centre power supplies: the grid will naturally produce good quality sine waves on the power lines, it gets distorted because non-linear (i.e. switched) loads like inverters get added onto it. That's why that data centre had problems - so the obvious solution is to modify the power supply design (potentially using this technology) to remove the problem at source and in the process get volumes up to push the costs down.

 

[peakbagger]

I see a lot more value in a concept that I have seen which has been referred to as packet power. Some other company owns the trademark to "Packet Power" so I don't know the actual name for the concept but here is how its supposed to work. A business or private individual has a smart appliance like a water heater. When the heaters controls determine that the heater is ramping off setpoint, the controls communicate to the grid and buys a "packet" of power. Preprogrammed into the units controls is priority, if the power is needed right now for something like a light bulb, the packet of power works like the current grid where its delivered immediately, the other option is the set the packet setting as a lower priority to only buy power when its cheaper. There already is a coarse method of this called off peak power where the appliance only runs during off peak and some utilities in CA have up to 5 tiered rates that vary over the day depending on demand. The packet concept extends the tiered concept to second by second and it automates the decision.

Using the water heater approach, the logic is established that if the heater temp is dropping but still within a band of upper and lower setpoints, the price the power user is willing to pay for a discrete packet of power is far less than when the heater has gone below the low setpoint. The owner can also elect to just set a low purchase power rate and put up with cold water if there is sustained demand. The customer is not dealing directly with the power company they are dealing witha middleman that combines the demands and contracts with the power suppliers. The middleman tracks the grid on near a instantaneous basis and the grid operator has variable instantaneous rate that is pegged to instantaneous power demand, if the system starts to detect slack where the amount generated is slightly exceeding the demand the instantaneous power rate drops and the middleman contacts the water heater to click on at discounted rate, once the system gets balanced the rate goes up until it exceeds the threshold programmed into the heater. If the supply and demand gets imbalanced in the short or long term, the power rate starts to go up and the middleman automatically taper off demand as the individual loads exceed their setpoint thresholds. If instituted system wide that keep the grid far more stable and reduces the overall system demand and transmissions losses as the transmission losses increase roughly with the square of the current being moved around.

This approach also incentivizes the middlemen to provide incentives to the customers to get them to agree to buy low priority power, it also incentivizes them to possibly put in fast response storage where they can elect put power back on the grid from storage when the rate is rising to and possibly delay low priority customer loads.

Technically with cheap fast communications on the internet its possible but only if the customers have smart appliances and the regional grid operators create an instantaneous market with protections in place to prevent fraud as once the middleman get big enough they can pull an Enron and artificially create systems imbalances by turn the individual demands on an of at opportune times. The other issue is businesses, high end homeowners and high end apartments would be the typical people to take advantage of the potential to save money while the poor homeowner and typical renter would not. They would end up paying a higher rate unless subsidies were put in place to upgrade their appliances.

The Tesla Powerwall has this capability built in to be remotely dispatched and that's why they want to lease them rather than sell them outright so they retain the right to implement this feature. They can borrow a homeowners power by telling the battery to stop charging and start discharging into the grid when there is short term imbalance on the grid. The owner is oblivious to this but Tesla gets an ongoing revenue stream by getting paid to have dispatchable instantaneous power available for the right price. Of course if the battery get sucked down and the grid goes off line the homeowner may be unhappy so Tesla can charge them a optional premium to keep the battery above a state of charge.