Hacker News Reader

I was with you until you said that distributed energy is more efficient than large scale energy.

A large scale hydroelectric, battery, or wind plant will be far more efficient than distributed energy in the home. Solar may benefit from capitalizing on under utilized residential real estate, but absent that incentive, a large solar plant will also be more efficient than a distributed per home grid due to standardized install and maintenance, larger inverters, and optimized layout (most home systems are non tracking).

You're not counting transmission loss, something that distributed systems avoid.

Central systems are more efficient but also must lose efficiency to transmit the power over distance.

As a naive layman, I'd be very curious to see numbers on inefficiencies of non-tracking panels vs. typical transmission loss.

25-35% more efficient for single axis tracking panels vs. non tracking. Distribution losses are ~6%.

http://en.wikipedia.org/wiki/Solar_tracker#Basic_concept

http://energyinformative.org/solar-panel-tracking-systems

I'm ignoring it because it's mostly irrelevant (6% total, ~1% for transmission). If it weren't, we would not have an electric grid.

http://www.eia.gov/tools/faqs/faq.cfm?id=105&t=3

https://en.wikipedia.org/wiki/Electric_power_transmission#Lo...

You are also missing the cost of the distribution and transmission infrastructure, which distributed generation avoids in addition to the transmission losses

Distributed generation using solar or wind without distributed storage actually needs more transmission infrastructure than centralized generation.

That's because of unpredictability of these power sources.

With centralized generation you can do hierarchical network, with smaller capacity further "downstream".

With distributed you need P2P network, and when some part of country has no sun and no wind - you need to be able to get that energy from other sources.

Distributed storage changes the game, but it depends on the costs.

You are missing anything that I said or linked to.

6% includes ALL losses for the entire distribution infrastructure. I am explicitly not missing that.

You also have not cited any evidence supporting your claim, so I don't know why I bother.

>> should they as electricity rates at night are dirt cheap

I dont know where you are but my electric rate is the same night or not... I pay the exact same rate 24/7

I do not have demand based pricing which is reserved for business customers or persons consuming a crap load more power than I.

>>At the end of the day

the Power Companies are doing their level best to penalize home use Solar, they want to charge home owners through the nose if they "sell" power back to the grid eliminating any costs savings and in come cases making a roof top solar project more expensive than just buy power from the mafia err power companies.

Distributed non-grid power should be the goal, Grid power needs to be phased out.

That's not necessary malicious on the part of the power company.

At a previous job, we built a system for a company that prepared batteries through multiple charge/discharge phases. During a discharge, the power was pushed back onto the grid.

The major challenge was matching the waveform of the grid power (occasional spikes confusing the zero-crossing, drift on the timer elements etc.). If you don't, it reduces the efficiency of the power company's transformer leading to replacement with a larger one, or damage due to the excess heat. You can imagine they're going to recover that expense from somewhere (and you and your dirty sine wave are pretty clearly at fault).

I would expect the home systems to be less precise than our industrial client's equipment.

A lot of UK homes have storage heaters that are powered at a cheaper rate at night on an off-peak circuit.

I never got the point of storage heaters. They're completely useless in the evening when you want heating. Much of the heat escapes when they're on overnight (while in bed so little need for it) and the rest during the day while you're at work.

The secret is good insulation and thermal design.

If the house temperature doesn't fluctuate wildly day to night, your scenario doesn't apply.

Plus you need a lot less heating.

As I understand it the big problem is the "duck curve," the spike in demand that occurs immediately as the sun sets when there's a lot of solar. If the batteries could smooth that out, that'd be pretty huge in helping solar adoption.

This is also the missing piece to move over to a more renewable energy source. Coal is on demand. If there is a need for more, you can spin up more factories, and many site waiting just for this.

Most renewable energies are reliable, but over longer periods of time. With a way to store energy, and have it dormant in the grid waiting to be used, this creates the surplus needed so we can switch to solar and wind and be able to rely on output.

Your point is correct, but the details are wrong: coal (and nuclear) provide so-called "baseload" power, while natgas and hydro provide "peak" power (this is why natgas plants are called peakers). Coal has a very long spin-up (and spin-down) time. If you've studied process control, the first-order time delay is on the order of hours to days. Natural gas and hydro have time constants of ca. minutes. Moreover to your point, batteries have very short time constants (but a plethora of of other problems).

So, for coal, is it just because of thermal inertia? The size of the whole thing? Do they need to load the coal in a specific way?

That's very likely due to size. Things like powerplants get more efficient as they get larger. (And as they get more complex by piling on more optimizing subsystems.)

Thermal inertia. Boilers have to be brought up to temp, and do not take well to throttling like a combined cycle natural gas plant.

For coal, I think it is because of the large turbines used. They have to be brought up to temperature gradually because they are very large precision made hunks of metal.

As has been mentioned, the boilers themselves do not heat up that quickly; however, you _also_ have to mix the pulverized coal with hot air and then pump it into the firebox for combustion.

If you're coming up from a cold start, you're going to need to pre-heat the air, probably with electricity and/or natural gas. With this in mind, it's not so much that they don't generate power right away, it's that they don't generate enough power to overcome the cost of their initial inputs.

From a power plant perspective, you aren't "started" until your output power is larger than your input power.

Plus for Brown Coal, you must run dryers powered by the waste heat. These take DAYS to get up to speed.

This is the answer. Commercial power penalizes you for having a large difference between your peak use and continuous use. If you can eliminate peaks using batteries you can eliminate this penalty. You can also recharge the batteries during off hours to cut costs further.

Unless you have a particular reason to worry about your connection to the grid, it's better to use demand management than batteries.

Let's take a Tesla S battery. It costs $30,000 [1]. It has a capacity of 85KWh, which lasts 265 miles. [2] The battery is down to 80% life after 100K miles [3], which is 377 cycles.

Say you fill up the battery at 5 cents/KWh and sell it at 20 cents/KWh. You've made $12.75. After 377 cycles, you've made almost $5000. You still have life in those batteries, but they are going to keep on getting worse, and the efficiency of energy in/energy out will start falling as well. Once you hit 40% they are probably useless. You might be able to run that $5000 two or three more times before you hit that.

You could only barely use the batteries to prolong their life. But you spent $30,000 on them! That's a lot of capital costs for something that isn't doing anything. And Li-Ions will gradually lose capacity even if you don't use them and store them in ideal conditions.

There are certainly efficiencies of switching from a car-bound battery to a stationary battery, although a lot of people posting on this page are talking about hooking up a car to do price arbitrage.

[1] http://my.teslamotors.com/it_CH/forum/forums/battery-replace... and watch how hard it is for the guy to get an answer

[2] https://www.google.com/search?q=tesla+car+battery+capacity&o...

[3] http://www.plugincars.com/tesla-roadster-battery-life-study-... don't just read the headline

Does anyone want to comment on flow batteries?

E.g. http://www.cbc.ca/news/technology/organic-battery-hailed-as-...

I assume the primary issue now is cost, but if you can decouple the reaction -> electricity engine from the reactants to a greater degree, then it sounds attractive? I imagine increasing storage tankage is a much easier problem to solve than increasing energy density.

I would expect the car is a much more harsh environment for the battery than my house. I'm not talking about vibration or temperature, though those matter, but electrical demand. A model S has a 310kW motor. Nothing in my house comes anywhere close to demanding 310kW. Of course the Tesla will rarely require the full 310kW, but its charge and recharge cycles are at the mercy of your driving.

You could smooth out the peaks in my home demand with probably just 10kW. Perhaps smooth out many of my neighbours peaks with 50kW. But my point is that there are no hard bounds that the battery needs to satisfy, unlike in a car when you have to supply the drive current or sink the brake current. You can tune a home battery's charge/discard cycles for maximum battery life rather than maximum performance. I don't have data for how much difference that would make to battery life, but I would guess it may be significant.

if the battery has capacity of 85 kWh and you're saying the motor consumes 310kW at max, then that would mean a model S only has a range (time) of 16 minutes?

Tesla estimates about 300 miles per charge at 60 MPH (300/60 = 5 hours)

http://my.teslamotors.com/goelectric#range

so if you get 5 hours out of 85 kw, that means you're using about 17 kW per hour

You don't go 60MPH when your motor produces its peak power.

Every time you accelerate, you're drawing way more current from the battery than you do cruising at highway speed. So the average power draw is 17 kW/hr, but some of that happens in great bursts.

To clarify, the P85D model has a peak draw of 490-510.

Using batteries seems like an insanely cost and environmentally-unfriendly way to engineer around billing practices of electrical utilities. I think we'd be much better off trying to get the numbers up for solar (installations, efficiency, lifespan, manufacturing) vs using vast quantities of highly toxic batteries powered from the grid from traditional power generation sources to game electrical utility billing for profit.

I really don't understand why this idea seems so popular with e.g. politicians.

I'm not sure that it is a safe assumption that this stationary battery will be based on lithium-ion cells. In the past, simple lead-acid batteries have far outperformed lithium-ion batteries in situations where weight and volume were not an issue.

Well thought-out post, however, your math is singularly reliant on the 30,000$ price tag for the battery and although you did list the source, it's accuracy is in some doubt. First of all, you can pre-buy a replacement Tesla battery for 12,000$(source: found in your same link [1]), secondly, the information is old.

I don't think there is any doubt that the number has come down significant. I'm sure you are aware that Tesla is building out a huge battery plant, significantly increasing lithium batter production ability, this will absolutely bring batter prices down.

I would be extremely surprised if their own internal numbers show a pay-off time of longer than 3 to 5 years. It's not like Tesla doesn't have people who get into the numbers, as the numbers are the only thing that is going to sell.

Without weight constraints of a car, you can use larger fire-containment-cell sizes, significantly lowering manufacturing costs. Also, you are ignoring the value you can get back when you recycle the battery, which may be part of how Tesla is offing the $15,000 replacement.

I still don't know if the numbers would work out though.

Wholesale LMPs are often more like $80/MWh at peak and -$60/MWh at nega-peak, but that still works out to 14¢/kWh, which is about the same difference you're calculating with.

Yep ... This idea has been around for a long time (Hydrokinetix was building ETS systems in the '80s) but a great battery will allow the technology to go mainstream. It's cool to hear that you've got such a backlog of orders!

Good luck!

Do you operate in one particular ISO or multiple?

smooth.