I'm about to move into a new house with a large damp/humid underfloor storage area. I want to force ventilate the area with a large 1/2 HP fan. To save a little electricity I plan to install 2 humidity sensors one inside and one outside and power the fan only when the humidity difference exceeds a threshold.
With wifi sensors and wifi power switches cheaply available I would like to control everything using the house wifi network. My son recommends using something like the Samsum SmartThings Hub but I don't want to spend >100 pounds only to discover the Hub isn't smart enough. Any experience out there?
Experience tells me that your damp, humid underfloor storage area could be damp and humid, possibly for days at a time, when we have foggy or wet weather.
The temperature of the air inside and outside also matters. Warm air can have a lower relative humidity than colder air while at the same time holding more water. so any control system may (will?) need to factor the temperature difference in as well as the relative humidity.
A friend had a similar problem and during a spell of warm dry weather he ventilated his cold damp basement and found it getting wetter! this counter-intuitive effect is apparently because the warm outside air, which seems dry, can often contain more water as vapour than the same air can contain when it is ducted into a cold basement, so the surplus water condenses out.
Rod makes a good point about the need to measure temperature as well, but once humidity and temperature are known it's not difficult to calculate the dew point and fire up the fan appropriately.
I'd use an Arduino to control the motor and a Raspberry Pi to provide network Connectivity. The RaspberryPi series are real computers, compatible with home networks. The 3B has built-in wifi, but if this doesn't work from a cellar, they also have a physical port which could be plugged into an Ethernet over Mains adaptor. They can run headless, ie no need for a keyboard or screen.
The Arduino Fan Controller part is rather simple: read the sensors, do some sums, and switch the fan. Also simple for the Arduino report status to the RaspberryPi and make it available over wifi.
Rather harder to send commands to the Pi, and then control the Arduino. The problem isn't connectivity or computer capability, it's the need to decide exactly what control is required and how it should interact with the user. Then the design has to be implemented in software. Human Machine Interfaces are generally the most difficult and complex part of the job. Part of the problem is users are almost completely unable to specify what they want in advance. They don't understand sequences, events, error inputs, ergonomics, or the need to back out when mistakes are made. The programmer needs a specification and gets a load of vague waffle! The same users are excellent at pointing out what's wrong with what they asked for when they see the finished result! Anyway, this aspect of the design needs careful consideration, and may need a few prototype attempts. Not rocket-science but time-consuming, especially if the requirement is blurred at the outset.
Best way of getting rid of water from a damp cellar is another question. In the good old days, a sump with a drain and pump plus natural ventilation, maybe supported a fire, was the only option. Before electricity automatic pumps were worked by water pressure from the supply.
Ventilation is a way of removing moisture and I think an intelligent fan would be effective. That the fan expels water from the building is an advantage compared with an ordinary humidifier, which has to be emptied.
How much water is involved may decide the issue. If the cellar is damp enough to fill a sump, then my first step would be to empty the sump into an external drain, probably with an automatic pump. The output of an air-drying dehumidifier could drain into the sump. If the cellar isn't damp enough to fill a sump, then replacing damp air with drier air from outside is a viable strategy, and I suspect cheaper than running a dehumidifier. But because ventilation is vulnerable to damp weather, a humidifier might be required as well. And of course there's a point where the best strategy is to seal the cellar to keep water out completely – pricey.
I would try a dehumidifier with a central heating condensate pump to get rid of the collected moisture. Changing the air might move some water vapour out but will also bring some water vapour in. Two steps forward one step back.
If you feel like making use of your Arduino knowledge then Indestructibles have a Dewpoint controller which is the best way of monitoring your conditions.
All good stuff, especially the comments about dew point.
1/2hp seems far more than you need, work out the capacity needed from changes per hour, 3 or 4 will do, more is for cooking or puddles of groundwater etc.
I have bought one of these for fume extraction, I still need to get it installed pulling through filters so I won't over reccommend it until it proves itself. Its worth noting that the biggest is only 120W. There is a size calculator on the page to the right.
Head over the parapet time because I'm terrible at maths!
How long would it take a ½ HP motor to empty a 3 x 4 x 2.5 metre cellar assuming a lift of 3 metres at standard temperature and pressure?
This is my answer in SMath Studio. I tried to use Units with odd results (what's a m³s) but believe the numbers are right:
Three seconds assumes the fan is 100% efficient, runs at full speed instantly and that that the outlet isn't constricted. 15 seconds might be more reasonable in the real world.
Replacing the air in the cellar quickly is an advantage. The purpose is to remove moisture, so faster the better; a 35W fan would take ten times longer, more if the cellar is wetting the air continually.
On the other hand, 35W fans are quiet. A big fan firing up in the cellar at 4am might wake the neighbours!
Happy to be told the calculation is wrong, and what I did wrong with the units!
Enthusiastic supporters of the Imperial System are invited to do the calculation in feet, pounds and seconds. One cubic inch of air weighs 0.00075oz. Extra points for answers in gallons.
Colin has only stated “underfloor storage area”, not “cellar”. Perhaps he might tell us more before anyone starts giving specific advice – which may be totally irrelevant.
He uses the word “area” too. Doesn’t seem to me to be a specific volume that is involved.
Rod makes a good point about the need to measure temperature as well, but once humidity and temperature are known it's not difficult to calculate the dew point and fire up the fan appropriately.
I'd use an Arduino to control the motor and a Raspberry Pi to provide network Connectivity. The RaspberryPi series are real computers, compatible with home networks. The 3B has built-in wifi, but if this doesn't work from a cellar, they also have a physical port which could be plugged into an Ethernet over Mains adaptor. They can run headless, ie no need for a keyboard or screen.
The Arduino Fan Controller part is rather simple: read the sensors, do some sums, and switch the fan. Also simple for the Arduino report status to the RaspberryPi and make it available over wifi.
Rather harder to send commands to the Pi, and then control the Arduino. The problem isn't connectivity or computer capability, it's the need to decide exactly what control is required and how it should interact with the user. Then the design has to be implemented in software. Human Machine Interfaces are generally the most difficult and complex part of the job. Part of the problem is users are almost completely unable to specify what they want in advance. They don't understand sequences, events, error inputs, ergonomics, or the need to back out when mistakes are made. The programmer needs a specification and gets a load of vague waffle! The same users are excellent at pointing out what's wrong with what they asked for when they see the finished result! Anyway, this aspect of the design needs careful consideration, and may need a few prototype attempts. Not rocket-science but time-consuming, especially if the requirement is blurred at the outset
Surely it’s far easier and less programming to connect arduino to router and not use raspberry pi.
Maybe the sketch for an Arduino to connect to the net and control a relay while reading a few sensors will be to big for it – as useful as they are they do have limitations ! Maybe an ESP32 for this ?
If you just want to compare the outside humidity to the inside and then control a fan an arduino is all you need but if you want to be able to read these off a smart phone or computer that is a different matter .
I personally would be more worried where the moisture is coming from and how to stop it getting in rather than removing it once it is there .You may need to look at some other technologies like sealing etc.
P.S
Dave the moisture control units i built using your modified sketch have now done two years service – with zero problems and zero condensation on my machines – probably cost me around $50 au a year to run if that !
Surely it’s far easier and less programming to connect arduino to router and not use raspberry pi.
Maurice
It's certainly an alternative, not necessarily easier. Arduino Boards don't come with Ethernet or Wifi, but a shield like this or this could be plugged in and programmed. It's the 'and programmed' part that's the challenge. I accept someone versed in Arduino's Ethernet library who knew nothing about Linux might well find it easier to buy a shield, program it and cut out the Raspberry. Particularly effective approach if the interaction is very simple.
RaspberryPi being a real computer means it supports Serial USB, Ethernet and Wifi at full throttle – all the tools and hardware are available. As such it makes a good bridge between the outside world and a microcontroller that only understands Serial I/O. Basically, the Raspberry does all the hard network stuff like DDNS and there's no significant limit to the size or complexity of the programs it can host.
As an example, I'm using a Raspberry as a Bridge to my Arduino clock at the moment. A short script writes whatever the Arduino is sending via USB to a file (which could grow to 5 or 6 Gb). As the raspberry links automatically to my home network, I can log in and view the growing file at any time. I can copy it to another computer without interrupting the capture, stop and start the capture, process the file on the raspberry, check how big it is and anything else file related. Apart from the script all the software is standard functionality. I don't have to program anything, but could if I wanted. Unlike the Arduino, which is tiny, the Raspberry isn't going to run out of puff.
Lots of answers and not a few requests for a better problem definition. So here goes …
The location is Phuket.The building is a bungalow on a sloping site. On the downslope side of the building there is a big enough space to just about stand up in.
The primary space is approx. 4-1/2m by 6m by 1.8m with one end underground and one end above ground. It contains a buried 5m3 spherical water tank with a loose fitting lid. This is thought to contibute to the elevated humidity. A secondary room contains the swimming pool pump and filter and is also very humid.
The fan is to be installed to vent the primary room into a secondary room. Both rooms have louvred doors to the outside.
Even now during the South Western Monsoon most days have dry periods and sunshine as well as torrential downpours.
A dehumidifier needs the room to be sealed and (I suspect) the water tank to have a more professional lid. A fan is cheap and cheerful and should improve the situation but I would prefer it not to run continuously at just under 1 GBP/day.
The temperature difference between the target room and outside is negligible unless the sun is shining in late afternoon and there is no wind and then we may get as much as 5 degC (I guess).
Measuring temperature is important for calculating the relative humdity. However, theoretically I need the absolute humdity level to decide whether to start pumping air. And the temperature inside and out will be approx the same and only varies from 25 degC on a cool night to 35 degC on a hot day. So basic humidity is all I need.
The fan needs to be mounted in a wall. Therefore I've bought something like this in order to be rugged and have a reasonable life span. Theoretically it will take around 15 seconds to evacuate the rooms volume of air. Yes I could have gone smaller but this was on a special offer at around 70 GBP.
I have some sonoff wifi switches looking for an application and I could easily use them here but the automation using humdity sensors while easy peasy with any programming language does not seem to be possible with commercial IOT automation devices. I have asked Samsung whether their SmartThing Hub can do what I want but they seem to be trying to avoid giving me any answer because of Covid!
A quick look at some ventaxia extactors on screwfix shows a model with 300-485m^3 extraction at 40watts or roughly 1KWH per day for 5 airchanges per hour. So running cost shouldn't be the issue.
…"Measuring temperature is important for calculating the relative humdity. However, theoretically I need the absolute humdity level to decide whether to start pumping air. And the temperature inside and out will be approx the same and only varies from 25 degC on a cool night to 35 degC on a hot day. So basic humidity is all I need."…
If you need a humidity sensor device, then I can confirm that is what I am using in the form of a Sonoff TH10 (plus the sensor).
The eWeLink app allows me to select temperature or humidity. I have it set to switch on my bathroom fan when humidity reaches 75%. and off when below this figure.
It's been working quite reliably for over a year now and still functions without a network connection once it has been programmed.
When as an apprentice at Vickers Weybridge, the high speed wind tunnel (Mach 6 !!) needed dry air. Fresh air was slowly passed over trays of silica gel and stored in an old barage baloon. When the tunnel ran, it consumed the dry air from the barage baloon – very quickly. Usual run time of the tunnel was about 6 seconds and it was just about possible to do 2 runs per day. Each run allowed the model to be moved 2 or 3 times (Quickly) and a schleron photo taken of the airstream passing over the model. I digress,
I was assuming the space was a cellar and the air had to be lifted out. Less work done by the fan shifting air sideways from an under-croft, but the principle is the same: how much air will be shifted in a given time by a motor of a certain power. I don't think fresh air coming in alters the sums because the replacement air does no work.
An equivalent way of ventilating the space is with a fire.
The fire and chimney are a heat engine that lifts damp air up the chimney which is replaced by fresh-air at the undercroft's entrance. As the heat engine obeys the same physical laws as the electric fan it's possible to calculate how much fuel (energy) is needed to remove a volume of air in a given time. Incoming fresh air doesn't alter the amount of work done by the fire either. I hope!
Dave, the incoming air does work against atmospheric pressure, which is in balance, so the outgoing air floats in the in-coming. Its more like moving a counter-weighted lift than moving the load on its own.
Its really a viscous flow problem, not a potential energy one. It would be interesting to know if your answer is anywhere near correct, it could make a good rule of thumb.
Great point Speedy Builder 5. The problem of moisture-laden air entering cold properties is a common one in historic monuments and can cause mould. The solution of drying incoming air (Air conditioning) is getting popular where budgets allow.
My son assures me that the Samsung SmartThings Hub will let me program my needs with user defined variables and loops and if thens so that's the way I'm going with wifi sensors and a wifi power switch.
When I get into the new house and have this automatic fan running I'll try to remember to post an addendum here. (If it's not running then it's a lot less likely that I'll post.)
Glad you're sorted out Colin. I'm still in disarray due to Roger's comments. Yesterday I proved conclusively to myself that Roger is wrong, only to realise my counter-argument was fallacious. Too many things making my brain hurt at the moment!
Dave
Author
Posts
Viewing 25 posts - 1 through 25 (of 25 total)
Please log in to reply to this topic. Registering is free and easy using the links on the menu at the top of this page.