Smarter Lighting with motion activation and target kWh per day?

[gernby]

Hi, I've been a member of this forum for a long time, and have been a marine aquarist for decades. However, I haven't been active on the forum much at all...

After years of neglect, I've started trying to bring my aquarium back up to its former glory. The aquarium and inhabitants are improving, but I'm trying to take it to the next level by applying some IoT expertise, and figured I would benefit from some input from you guys on some of it.

In short, I work from home, and like to have the aquarium lights on about 14 hours per day. Also, since I'm an engineer with a habit of overdoing things, my total LED power potential is above 600 watts. I'm using a mixture of cool white, royal blue, and several flavors of violet with 6 large heat sinks from RapidLED. I've also added a second aquarium that's half the size of my first, and am running it in parallel on the same filtration system.

My 180g extreme bowfront has 4 fixtures that can put out just over 400 watts, and my 90g bowfront has 2 fixtures that can put out just over 200 watts. I love the way it looks when the lights are at full power, but it's way too much light to run for the number of hours I want to run them.

So what I've done is to link the tank lights to my Phillips Hue motion sensors (via Node-Red), so that I can dim the lights (slowly) whenever no motion is detected in the room, and brighten them (quickly) whenever I walk into the room. It works extremely well, and doesn't seem to bother the inhabitants. It also seems to be reducing the algae problem I've been dealing with.

Anyway, it occurred to me that there might be some "ideal" daily kWh that I should be targeting, and just alter the distribution of those kWh's based on the motion sensors. On days that I walk through the room a lot, the lights would consumer power earlier in the day, but on days that I'm gone, the lights would consume most of the power late in the day.

It seems that using this approach might allow me to dial in an ideal amount of daily "lightage" that works best for for the inhabitants, while also getting the best visual experience.

Is this something anyone has tried? Thoughts about it? Any guesses about how many kWh / gallon I should target each day? I already have power monitoring on the circuit, so I can see instantaneous and cumulative power usage in Node-Red. I've been averaging a bit over 5 kWh per day recently (EDIT: down to 4 kWh today). When no motion is detected in the room, the total LED power output is about 250W, and when motion is detected, it climbs to 350W for at least 3 minutes. The descent from 350 to 250 is much slower than the climb.

[gernby]

I’m hesitant to post this, since I know it’s just begging for criticism about how crappy everything looks...

However, it was MUCH worse a few months ago...

[outy]

No big deal starting over can be a good thing.

With so little coral you are wasting wattage though.

I have 3 fixtures over 215g running about 225 W total measured by a kill a watt meter.

Tank is loaded with coral and all are happy.


You need to mimic their natural habitat to some extent. I do B 11-9 and W 12-8

[oreo57]

Well a bit of a complicated approach and based on the daily light integral (DLI) approach to photosynthesis.

First thing you would need to do is estimate PAR/Kw
Then you would need to know the DLI your tanks like..
After that it's just a matter of adding "periods" to equal your target DLI..

Sounds simple huh..


so say you run a constant PAR of 300 for 8 hours.
The instantaneous wattage is 300 for the lights, as measured.

Running your lights for 16 hours at 150w would be an equivalent exposure..
(same kWh)
Obviously ramping makes the math tricky for all but the finest nerds..

"The formula for DLI is simple.(LOL) It is Photosynthetically Active Radiation (PAR, reported as micromole per square meter per second, or molm²sec) times the photoperiod in seconds

Of course you would just substitute kW eventually... micromoles/kw
and calculate total for one day.
Varying "time periods" is another variable.
Pretty sure it can be done.. but the complexity is huge ..for most..
Of course that is what computers/programs/ and data loggers are for..
I'm sure somewhere its done all the time.
Maybe different parameters, but same exact concept.

http://.com/threads/daily-light-inte...ussion.312308/
(full link not allowed)

Any guesses about how many kWh / gallon

Old metric.. need PAR to be effective..

[gernby]

Originally Posted by outy

No big deal starting over can be a good thing.

With so little coral you are wasting wattage though.

I have 3 fixtures over 215g running about 225 W total measured by a kill a watt meter.

Tank is loaded with coral and all are happy.


You need to mimic their natural habitat to some extent. I do B 11-9 and W 12-8

Wow, that's good to know that such a large tank of corals can do well with just 225W! Any idea what LED's are in your fixtures?

I don't really understand "I do B 11-9 and W 12-8". Does that mean you do only blue from 11 PM - 9 AM, then run blue AND white from 12 PM - 8 PM?

[gernby]

Originally Posted by oreo57

Well a bit of a complicated approach and based on the daily light integral (DLI) approach to photosynthesis.

First thing you would need to do is estimate PAR/Kw
Then you would need to know the DLI your tanks like..
After that it's just a matter of adding "periods" to equal your target DLI..

Sounds simple huh..


so say you run a constant PAR of 300 for 8 hours.
The instantaneous wattage is 300 for the lights, as measured.

Running your lights for 16 hours at 150w would be an equivalent exposure..
(same kWh)
Obviously ramping makes the math tricky for all but the finest nerds..

Of course you would just substitute kW eventually... micromoles/kw
and calculate total for one day.
Varying "time periods" is another variable.
Pretty sure it can be done.. but the complexity is huge ..for most..
Of course that is what computers/programs/ and data loggers are for..
I'm sure somewhere its done all the time.
Maybe different parameters, but same exact concept.

http://.com/threads/daily-light-inte...ussion.312308/
(full link not allowed)

Old metric.. need PAR to be effective..

Thanks so much for your input! Lots of good info to consider!

I understand that the only way to calculate optimal kWh / day for my system would require PAR values, but since A) I don't have a PAR meter, and B) I don't know what PAR levels my corals would like, it seems that a guess based on kWh/day might be the best starting point, then adjust target kWh/day over time based on what the aquarium seems to like.

Although, it does seem that I should verify that PAR per Watt is linear across the range that I intend to use. I'm not certain that PAR at 200W would be exactly half what it would be at 400W. I'm using Meanwell LDDH drivers with PWM dimming on an HEP-600-54 power supply, if that helps.

The algorithm would actually be pretty simple. If the LED watts bounce between 200W and 300W based on motion detection, and the lights will be on for 14 hours, then minimum kWh per day would be 2800 with 0 motion. If I determine that optimal kWh per day for my tank is 3200 kWh, then the system would need to run at the higher wattage for 4 hours each day.

In the case that no motion is detected throughout the whole day, then the system would need to increase light output from 200 to 300 during the last 4 hours of the day. However, if the system runs at the higher wattage for 1.5 hours due to motion detection earlier in the day, then it would only need to run at the higher wattage for the last 2.5 hours of the day. The system just needs to keep track of the cumulative motion activated minutes throughout the day.

Actually, since my lights are on a TP-Link HS110 smart plug, the system really wouldn't need to "book keep" the number of motion activated minutes at all. It would just need to periodically poll the smart plug for the current watts and current day's kWh, and adjust the "plan" accordingly.

The only way I think it would be complicated is if I tried to include a long "sunset" period at the end of the day.

[outy]

Originally Posted by gernby

Wow, that's good to know that such a large tank of corals can do well with just 225W! Any idea what LED's are in your fixtures?

I don't really understand "I do B 11-9 and W 12-8". Does that mean you do only blue from 11 PM - 9 AM, then run blue AND white from 12 PM - 8 PM?

Blue starts at 11 and stays on until 9 pm all lights off.

White starts and hour after lights on, then an hour before they all shut off

All leds grow coral fine. They are not cree

[gernby]

Originally Posted by outy

Blue starts at 11 and stays on until 9 pm all lights off.

White starts and hour after lights on, then an hour before they all shut off

All leds grow coral fine. They are not cree

Got it...

I was asking which LED's you have just to to see if your lumens per watt might be similar to mine. My aquariums are both very deep, and the corals are all at or near the bottom. I also don't use any lenses, so the light disperses outwardly a LOT.

I actually wonder if the lack of lenses might be contributing to my algae issues, but I've just had no luck getting lenses to work. I basically just ruin the lenses and LEDs trying to get them to stick.

[gernby]

I got it working well, and the inhabitants really seem to like it! I have the lights start coming on very slowly at 9:00 AM, then follow a half sine wave until they slowly turn off near midnight. The peak is around 4:20 PM, but I also have the system linked with my Pillips Hue motion sensors, which trigger a temporary "bump" in output whenever motion is detected in the room. I really like the effect, since it's kind of like a break in the clouds.

An unexpected benefit is that I get to see a much broader set of behaviors from the inhabitants! The invertabrates and fish all act differently at various times of the day, especially the cleaner shrimp, which I rarely saw before.

Below is a partial snapshot of my home dashboard, which shows 2 full daily cycles. The lighting graph is the 3rd graph, which shows instantaneous wattage on the left and cumulative kilowatt hours on the right. The graph below that is similar, but shows usage from my 2 heaters. I turn the heaters off during "On-Peak" pricing from 6 AM-8 AM and 3 PM-8 PM. The upper 2 plots show total household power usage. The aquarium is the biggest consumer, by far... (during the winter).