Maximising Internal Use of Power

[DanS26]

Hoosier energy, the G&T for my REMC POCO, is working on a solar PPA for $0.035 kWh. Solar costs are coming down at the utility scale for sure. Hoosier's production cost for coal powered energy is ~0.030 per kWh.....so solar is coming close but its main drawback is it is intermittent.

My costs are low because of no labor costs included.....if labor was included I estimate my LCOE would be around $0.06-0.07 per kWh. Still less than average but I also was able to purchase BOS parts at contractor's price which reduced the cost of my ground mounts significantly.

If you're handy and can DIY these kind of projects are fun and profitable, but you need to read, understand and follow NEC guidelines.

[DanS26]

The longer the time frame, the lower the LCOE becomes. My assumptions are different than Dan's, but using a 12 yr. lifecycle and post incentive cost, my real LCOE is ~~$0.1274/kWh. If I stretch that out to 25 years, the LCOE drops to $0.073/kWh. Cash purchase. Zero salvage value, 6% discount rate, electric rate inflation = 3.28/yr. = general inflation.

Not arguing numbers here, just pointing out that the longer the lifecycle, usually, using common assumptions, the lower the LCOE.

If I really thought my system had only a 12 year lifecycle........there is no way I would have committed to the project.

[J.P.M.]

If I really thought my system had only a 12 year lifecycle........there is no way I would have committed to the project.

Opinions vary. I'll be dead, moved or age demented and in a padded cell playing fungo with my poop by then. Besides, and with zero salvage value, after 8 more years, on a 12 yr. lifecycle, technology will have surpassed what's on the roof now. If so, and a thing only worth what someone is willing to pay, maybe a 12 yr. old system will be viewed as not the latest/greatest just like a 12 yr. ond solar water heater that's perfectly functional being viewed as an obsolete eyesore. and giving negative price enhancement to a property or negative salvage value.

Makes no difference to me if someone uses a 50 yr. life cycle for LCOE calcs, or 6 months, or anything in between. Just pointing out that longer life cycles lower LCOE's. That's one of the tricks peddlers use to make things look cost effective. Most marks are clueless and see what they want to see and then parrot the payback the peddler feeds them.

With the average time in a house in the U.S. of ~ 7-8 yrs. and resale enhancement from PV being at best suspect or wishful thinking, I'm not sure how realistic a 25 yr. lifecycle is. I thought 12 yrs. to be a more realistic number, but that's just my opinion.

[DanS26]

OK fair enough and I can see your point. The analyst in me says that if a system has a realistic 25 year life then the LCOE should be based on 25 years whether I live to see that return or not.

The reason is so that I can make a valid economic decision on how I use the energy produced during the asset life cycle. If I don't know my true underlying costs on a per unit and apples to apples basis then its hard to make an economic decision on how I use that asset. Most people do not have the flexibility on energy sources so its a mute point, but I do, and I want to know if it is better to burn propane or heat ceramic bricks or heat water or charge my EV or sell wholesale back to POCO.

[J.P.M.]

OK fair enough and I can see your point. The analyst in me says that if a system has a realistic 25 year life then the LCOE should be based on 25 years whether I live to see that return or not.

The reason is so that I can make a valid economic decision on how I use the energy produced during the asset life cycle. If I don't know my true underlying costs on a per unit and apples to apples basis then its hard to make an economic decision on how I use that asset. Most people do not have the flexibility on energy sources so its a mute point, but I do, and I want to know if it is better to burn propane or heat ceramic bricks or heat water or charge my EV or sell wholesale back to POCO.

Understood. I'd take some question of the choice of 25 years as being or seeming rather arbitrary, but that's not a knock - just my opinion.

As a matter of fact, all this cost estimating and what amounts to engineering economics is more than a small part of what got me into all this R.E. stuff in the first place back in the '70's. I was living in Buffalo and bored with life as a peddler, trying to figure out how to stay reasonably warm in the winter without going broke. Then, my wife at the time, who was and still is an accountant, introduced me to the joys of life cycle costing. Later, as an engineer, and armed with engineering economics I needed to get involved in comparison of alternative analysis with a lot of financial considerations included in the decision making process.

[jwrgorman]

This sounds like a very cool DIY implementation to me. I am interested in both the optimisation of internal power use based on the forecasting of energy use - wrsnz are you using any methods of trying to prediction energy generation or usage and then acting on that forecast? I think it needs a feedback loop so your system would learn from mistakes - over/under guesses - and keep refining based on experience from empirical data. We are developing a module that aims to use a neural network (like Google TensorFlow for example) to train on experience and environmental conditions of when power was generated/consumed to produce energy forecasts. I see you're using Delphi - a cool OO dev environment from what I remember (I was psyched about Kylix when it came out!) we are using Java OSGi for our platform but it's an open-source project so happy to collaborate. The platform is quite developed but this new module (SolarQuant) is pretty new: http://solarnetwork.net/v4/solarquant/ and could probably train on your system's data to try to find good decision points. the idea is that it sits chugging away on a box (multi Intel or nVidia CUDA accelerated) or in the cloud and just gives advice via a RESTful API, you make the calls on hardware switching decisions and implementation yourself.