4 ms·

Actually people have analyzed the issue at scale and concluded it's doable with renewables. It's not that difficult to find these studies, if you actually care

by olau 5y ago

Actually people have analyzed the issue at scale and concluded it's doable with renewables. It's not that difficult to find these studies, if you actually care to look for them. Whether you believe those studies or not is immaterial, though, because the economics of renewables are going to win in the end. Ten years ago you could have reasonable discussions about whether that was the case or not, but the trends did continue. The other day I saw a forecast of 0.04 USD/kWh for offshore wind within a few years.

robomartin 5y ago
Care to provide at least one link for one of these studies? I have read many. I am not in the habit of opening my mouth unless I have done enough research to confirm my hypothesis. I can still be wrong, yet, based on what I have seen in the domain, most of these studies have an element of "then a miracle occurs" in them. This can take the form of highly theoretical --when in real life things are never linear-- or a bunch of hand-wavy stuff that can't support the conclusions reached. I made some adjustments to my models and was able to get the cost of solar with batteries down. However, to be fair, I now need to factor in the actual solar generation efficiency (meaning, panel rating vs. what you actually get at the peak) as well as to account for the fact that the actual output power curve is roughly an inverted parabola. The energy produced by the system, then, would be the integral of the power curve over the period of solar production (roughly 7:00 AM until 5:00 PM). One must then take into account that system efficiency (DC to AC conversion, battery charging, battery discharging) must consider a range of variables. For example, the Tesla Powerwall 2 is rated at something like 90% efficiency while only drawing 3.3 kW at what I would all laboratory-type environmental conditions: 25 degrees C. I am having trouble finding data on actual efficiency under real environmental conditions, hot and cold. Beyond that, the efficiency of the cells (for any battery, not just Tesla's) degrades over time. In Tesla's case, they guarantee 70% over 10 years. This means also modeling the replacement of millions of battery packs per year at or around the ten year lifespan range. Losing 30% of energy production would be a problem. So, you either overbuild or replace everything as soon as it drops below a certain threshold. Over the 20 year deployment model I created, 5% of the battery packs would have to be replaced every year starting somewhere around 7 to 10 years after installation. Then there's the lifespan and behavior of solar panels. I'll make the assumption that at the 20 year warranty expiration point, most solar panels will be at about 80% production. This assumes no other environmental factors are in play (dirt, scuffing of the glass, etc.). To finish the model one would have to include the negative temperature coefficient of panels: As the operating temperature increases, output decreases. A typical panel might have a coefficient of about -0.35% measured from a baseline of 25 degrees C. At 35 C, the panel produces 3.5% less energy. Panels can run very hot (60 C) because, well, the sun will heat them...they are not white reflectors. At 60 C the loss can be as high as 12%. This, BTW, is why I installed three large (~6 ft diameter) fans under my solar array. I haven't done the math yet but I suspect they pay pay off in efficiency during the hottest periods of operation. The model would have to allow for both failures and replacement due to degradation. I have friends with arrays consisting of around 15 panels who have already had to replace two or three panels after ten years. Once again, reality is far more stochastic than data sheet specifications. Anyhow, the point is that the real-world model of solar energy systems is far from the simple idea of a solar array delivering power consistently and at peak performance.
- orthecreedence 5y ago
  > most of these studies have an element of "then a miracle occurs" in them. I wonder how many of them have been updated to take into account constant wildfire smoke blotting out the sun in peak summer months. It seems insane to me that we keep doubling-down on weather-dependent energy generation when everything we know about the weather is now in flux. Especially when we have a perfectly safe, scalable, and practically zero-carbon way to generate power sitting right next to us collecting dust.
- robertlagrant 5y ago
  This is why I like HN. "My point of view is going to be borne out as being correct; if only you bothered to investigate you'd have it too." "Well, I have, and I've done it in practice, and I understand the maths, and here are my concerns."

by olau 5y ago

robomartin 5y ago

Care to provide at least one link for one of these studies? I have read many. I am not in the habit of opening my mouth unless I have done enough research to confirm my hypothesis. I can still be wrong, yet, based on what I have seen in the domain, most of these studies have an element of "then a miracle occurs" in them. This can take the form of highly theoretical --when in real life things are never linear-- or a bunch of hand-wavy stuff that can't support the conclusions reached. I made some adjustments to my models and was able to get the cost of solar with batteries down. However, to be fair, I now need to factor in the actual solar generation efficiency (meaning, panel rating vs. what you actually get at the peak) as well as to account for the fact that the actual output power curve is roughly an inverted parabola. The energy produced by the system, then, would be the integral of the power curve over the period of solar production (roughly 7:00 AM until 5:00 PM). One must then take into account that system efficiency (DC to AC conversion, battery charging, battery discharging) must consider a range of variables. For example, the Tesla Powerwall 2 is rated at something like 90% efficiency while only drawing 3.3 kW at what I would all laboratory-type environmental conditions: 25 degrees C. I am having trouble finding data on actual efficiency under real environmental conditions, hot and cold. Beyond that, the efficiency of the cells (for any battery, not just Tesla's) degrades over time. In Tesla's case, they guarantee 70% over 10 years. This means also modeling the replacement of millions of battery packs per year at or around the ten year lifespan range. Losing 30% of energy production would be a problem. So, you either overbuild or replace everything as soon as it drops below a certain threshold. Over the 20 year deployment model I created, 5% of the battery packs would have to be replaced every year starting somewhere around 7 to 10 years after installation. Then there's the lifespan and behavior of solar panels. I'll make the assumption that at the 20 year warranty expiration point, most solar panels will be at about 80% production. This assumes no other environmental factors are in play (dirt, scuffing of the glass, etc.). To finish the model one would have to include the negative temperature coefficient of panels: As the operating temperature increases, output decreases. A typical panel might have a coefficient of about -0.35% measured from a baseline of 25 degrees C. At 35 C, the panel produces 3.5% less energy. Panels can run very hot (60 C) because, well, the sun will heat them...they are not white reflectors. At 60 C the loss can be as high as 12%. This, BTW, is why I installed three large (~6 ft diameter) fans under my solar array. I haven't done the math yet but I suspect they pay pay off in efficiency during the hottest periods of operation. The model would have to allow for both failures and replacement due to degradation. I have friends with arrays consisting of around 15 panels who have already had to replace two or three panels after ten years. Once again, reality is far more stochastic than data sheet specifications. Anyhow, the point is that the real-world model of solar energy systems is far from the simple idea of a solar array delivering power consistently and at peak performance.

orthecreedence 5y ago
> most of these studies have an element of "then a miracle occurs" in them. I wonder how many of them have been updated to take into account constant wildfire smoke blotting out the sun in peak summer months. It seems insane to me that we keep doubling-down on weather-dependent energy generation when everything we know about the weather is now in flux. Especially when we have a perfectly safe, scalable, and practically zero-carbon way to generate power sitting right next to us collecting dust.
robertlagrant 5y ago
This is why I like HN. "My point of view is going to be borne out as being correct; if only you bothered to investigate you'd have it too." "Well, I have, and I've done it in practice, and I understand the maths, and here are my concerns."