Hacker News new | past | comments | ask | show | jobs | submit login

Residential-only energy use is irrelevant. What matters is peak grid usage (ie. residential+industrial) and it is actually mid-day—perfect for solar. For example the UK's real time grid monitoring system clearly shows this peak on the weekly chart: https://i.imgur.com/wTIt5tg.png (source: http://www.gridwatch.templar.co.uk/)



A weekly chart like that isn’t really granular enough to show that.

This chart is a few years old, but it shows that the peak demand was in the early evening:

https://www.eia.gov/todayinenergy/detail.php?id=830

The peak demand will also vary at different locations and times of year. As the wikipedia article states:

>...It depends on the demography, the economy, the weather, the climate, the season, the day of the week and other factors. For example, in industrialised regions of China or Germany, the peak demands mostly occur in day time, while solar photovoltaic system can help reduce it. However, in more service based economy such as Australia, the daily peak demands often occur in the late afternoon to early evening time (e.g. 4pm to 8pm).

https://en.wikipedia.org/wiki/Peak_demand


Your chart is even less granular than mine (hourly vs half-hourly), and it has a timezone conversion error. The source they use (ISO New England - https://www.iso-ne.com/isoexpress/) shows the peak of a typical October month to be between 3pm-4pm and not 6pm-7pm: https://i.imgur.com/5rWlnVO.png (the 3 hours difference is the timezone offset between east coast and west coast.)

My chart, if you look closely, has a granularity of one half hour. If you want something even more granular, ISO New England data more clearly shows the peak to be between 2pm-3pm for a typical summer month: https://i.imgur.com/rX1JqoJ.png


>Your chart is even less granular than mine (hourly vs half-hourly),

I think you might have posted the wrong image to imgur. You linked to: https://imgur.com/wTIt5tg

I am not sure how you are getting half hour indications on that weekly chart.

>...ISO New England data more clearly shows the peak to be between 2pm-3pm for a typical summer month:

Yes, that was why wikipedia says:

>...It depends on the demography, the economy, the weather, the climate, the season, the day of the week and other factors. For example, in industrialised regions of China or Germany, the peak demands mostly occur in day time, while solar photovoltaic system can help reduce it. However, in more service based economy such as Australia, the daily peak demands often occur in the late afternoon to early evening time (e.g. 4pm to 8pm).

https://en.wikipedia.org/wiki/Peak_demand

The EIA web site was talking about October, not the summer months when air conditioning would be used.

As another data point, With PG&E pricing, the most expensive hours are noon to six:

https://www.pge.com/en_US/business/rate-plans/rate-plans/tim...

I think my point is that statements like "hat matters is peak grid usage (ie. residential+industrial) and it is actually mid-day—perfect for solar." can over simplify the problem. The amount of demand will vary in different areas at different times of the year and even when it is raining, there will still be demand.


«I am not sure how you are getting half hour indications on that weekly chart.»

See https://i.myimg.io/wi9J.png

«The EIA web site was talking about October»

This is why I also showed an October chart, with a peak at 3-4pm.

But, yes, peak can vary with weather, time of year, etc. All I'm saying is that in general—not always—peak PV generation coincide reasonably with peak electricity demand.


>...See https://i.myimg.io/wi9J.png

OK, for many reasons we are just going to have to agree to disagree on that chart. (I just measured https://imgur.com/wTIt5tg and on my machine it definitely is more than 41 pixels between each day but that doesn't mean each pixel is representing exactly x minutes of time, there is no indication on the chart each pixel actually represents 35 minutes or if it is interpolated data, etc. To me it gives a rough showing of energy demand over a week.)

>...This is why I also showed an October chart, with a peak at 3-4pm.

It isn't clear to me why there would be peak demand at 3-4 in the afternoon in an October day. As the EIA site says:

>...During this period, usually in the early evening, operators need more generating capacity–including more costly "peaking" units. Both day-ahead and long-term forecasts account for these peaks to ensure the assignment of adequate capacity.

>...But, yes, peak can vary with weather, time of year, etc. All I'm saying is that in general—not always—peak PV generation coincide reasonably with peak electricity demand.

I think we basically agree on that. Peak PV generation is probably closer to noon, but in general more energy is needed during daylight hours than at night. (And obviously a hot sunny day will spike energy usage for cooling and solar PV is a perfect fit at that time.) The problem to me is when people over simplify this issue. The amount of demand will vary in different areas at different times of the year and even when it is raining, there will still be demand. I've always been a supporter of solar power but there is a lot of handwaving going on when people say we can get 100% of our power from solar and wind without there being some huge advances in energy storage.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: