Thought I’d update this again, spurred on by a post from @Ronski regarding getting sick of the heat…
Since the farm went off on the 23rd June, it has not yet come back online. The average forecasted highs have been pretty accurate, so had the farm auto-restarted, it would have shut off again a few hours later.
The only changes I have made to the system is the re-start temperatures. Now, if the average forecasted maximum daily temperatures of three close surrounding locations across the 5 days does not exceed 17 C and any of the location’s maximum forecasted temperatures across the 5 days does not exceed 19 C, it will power on the farm.
The previous setpoints of 20 and 22 C were too high, and would have resulted in the farm shutting down again shortly afterwards. I suspect the new 17/19 C figures will be about right.
It used to bother me that anything less than 100% uptime meant I was losing out, but now, with the price of XCH where it is, I can stomach the farm being down and not producing heat I don’t need…
Maybe those 20-22 C temps were fine, but you need one more condition - how long such temps were staying during a day. This way, you can discriminate between the summer (several long hours per day) and winter (just a one-off spike in only one day during the monitoring times). If you also add such duration in the shutting down the farm code, you could avoid turning off / on your farm during such spikes (maybe the fall and spring are full of those conditions).
.
.
By the way, I got three ESP32 boards (https://www.ebay.com/itm/176053628906) and several temp sensors (https://www.amazon.com/dp/B0B4D69MNB?ref=ppx_yo2ov_dt_b_fed_asin_title) to control the external fans, but kind of cannot motivate myself to finish the job.
Yes, I could. I would get the hourly forecast for this.
This is kind of already catered for as my cooling fan has to be running for more than a certain time (like 2+ hours continuously) along with a few other conditions, such as forecasts above a setpoint, before the farm shuts down.
Of course, I could do all manner of things, but given it cost as much to farm as it makes now, I shan’t bother any more until I need the small amount of background heating again…
The farm has remained on ever since, given that the weather has ‘changed’ to autumn mode in what appeared to be an almost overnight event, it’s not surprising.
At current status, it looks as if the farm would power off once a year in June/July and not power back on until September/October, weather and temperature dependant of course. Happy with that.
Next up on the to-do list. Pull kWh cost from the internet for my area, pull XCH price from the internet, and calculate in real-time whether I am profitable or not, and use this as a secondary means to determine when to power on and off the farm, with a bit of hysteresis of course.
Although at present, I think it would power off and not come back on at all!!
Of course, I would also factor in the cost of heating the office otherwise if the farm wasn’t there. There is only electricity available, unless portable gas heater was used…