Daytank system with some electronics.

[pmagowan]

In my design of a new 40 odd foot sailing boat I have been looking at various fuel and water tank possibilities. I have come to a few conclusions and wanted to run them past the panel.

I want to have reasonable capacity for fuel being that this will be a potential bluewater cruiser. The problem with fuel is weight so I will try to put the tanks as low as possible. To do this I will have to split them into two tanks, one on either side running along the hull. They will be integrated into the hull for most efficient use of space and also because they then provide a 'double-hull' for added protection from sinking. I will then need a 'day tank' for feeding the engine. This could be considerably smaller and above the engine for ease of bleeding and to prevent air sucking into the fuel lines.

All this will need to be controlled throught the transfer of fuel from tank to tank. I have just discovered microprocessors and controllers such as the raspberry pi and the arduino and was thinking that this could be useful. My idea is to have a fuel level sensor in each tank, an 'in' valve (motorised) and an 'out' valve for each tank and a pump with a racor filter on it. The tanks will be connected in parallel with the pump so that transfer of fuel from one tank to another simply involves opening the 'out' valve for the tank giving fuel and opening the 'in' valve for the tank recieving fuel and then turning the pump on.

This could all be controlled by a microcontroller which is in turn controlled by a Raspberry pi with a nice user interface (The pi could also handle GPS data, media, battery monitoring, engine monitoring etc). So, for instance, in normal operating mode the day tank would be kept full with the level being monitored and the fuel transfered as required. The tanks would be kept ballanced. There would be fail to safe, as in if the pi died the microcontrollers would default to normal operating and if they died there would be manual override and if the electrics died there would be a manual fuel lift pump.

I know it sounds a bit complicated but I would be interested in your thoughts as the system would have loads of plus points: fuel polishing, weight distribution, tank capacity, gravity feed to engine etc. I see the Pi as being a central hub on the boat which could distribute wifi, nmea data and multimedia as required as well as providing a nice user interface to see the level of tanks, temperatures of engine and batteries, leak detection etc etc.

 

[lw395]

Good luck with fixing that with your head in the bilge in the middle of the night.

It all sounds wrong on many levels.
For a start, I wouldn't have a tank each side, you'd have to fill them equally, so a dirty fill would go in both.

A day tank just needs a few float switches, to turn the pump on and hi/lo alarms.

 

[rob2]

My instinct was against it , but you recognise the need to fail safe and are proposing two levels of back-up so if you can be bothered, then why not? I disagree about contamination as you propose pumping through a filter to the day tank and presumably filtering again on supply to the engine. If you're carrying a lot of fuel and storing it for long distance cruising, it would be as well to use a filter/separator funnel when filling.

My main concern is how you are going to get a usable level indication to balance the fuel between tanks. As the tanks will be assymetrical, when heeled one will likely read high and the other low. I wonder whether a sonic level sensor would work when under way?

Rob.

 

[[70521]]

A few valves and a hand pump should do it.

 

[quimby]

Keel tank, if not lifting keel. with access plate for cleaning keel tank in the future. Pump to day tank manual and electric. via filters.

 

[pmagowan]

The way I look at it there are 3 main areas
1) the tank placement for best boat stability
2)the best arrangement for engine running
3)The best arrangement for ease of use

1) I am pretty sure that my current arrangement would be best for stability, reducing roll and keeping weight low. I am using other low areas for water tanks also. I am at very early stages in the design so I do not have all the calculations done so the numbers will need to be crunched when the computer has all the info. There is the potential for the tanks to be adjusted so that water was split and the fuel occupied the space directly over the keel but I would be more concerned about the size of a single fuel tank and problems with baffling etc. There is the potential to balance the boat by transferring fuel from one side to the other or even to empty the tanks for floatation.

2) I am pretty sure that a day tank system is the best for running the engine. The tanks would be too low down to be easily fed with the fuel lift pump and there is always the problem of bleeding and air with a tank-below-engine system. The day tank would only contain pre filtered fuel (through the bulk tanks) and this would be filtered again on the way to the engine.

3) My current boat has a single tank (about 40L) and this is topped up by jerry cans. Really the only difference here is joining all the jerry cans into larger tanks and then transferring them by pump rather than by pouring them in. So you can look at the system as a standard gravity fed one with a couple of bulk tanks to store extra fuel. Once you decide that this system is fine then it only matters how you use it. I could simply have a system where you go down below, open the tank and then pump it into the day tank (this is the equivalent of pouring in a jerry can). In fact, my system has this but then I also have a system 'above' it that simply does this automatically. I then have a further system that has an interface, i.e. a touchscreen where you can look at all the levels etc as you please and control the system as you please. The thing is that if you pay it no attention then it would just tick away in the background keeping the day tank topped up.

At the end of the day if anything fails it will fail down a level towards the default manual level which is where we start. I have never liked filling the tank from a jerry can especially in a bad sea when it is raining (even with a Racor funnel).

To answer other comments:

Lw395: you wouldn't need to fix it as worse case scenario you have a full day tank to last you for the day. Fuel would be polished in the system as it is, by default, also a fuel polishing system so there would be no problems with contaminated fuel. I don't see how this would be any different from the average setup anyway other than I could never get contaminated fuel into the day tank without it passing through both a Racor funnel and a Racor filter to the micron level.

Rob2: the fuel level is a difficulty but the tanks would be baffled and actually quite narrow and deep so less effected by heel. The other thing is that if the sensor is central it could be reasonably calibrated. You could go crazy and write a programme to calibrate the sensors to level of heel but I am not sure I need it any more accurate than a standard gauge. Fuel can not flow from the high tank to the low tank so it is only internal tank movement that needs accounted for.

Camelia: A few valves and a pump will do it. The micro controller is simply a switch to operate the valves automatically.

Quimby: I have mentioned above regards the setup but it could be that water and fuel tanks could be swapped so that the keel tank would be fuel.

 

[lw395]

Many people use the phrase 'day tank' to mean a fairly small tank.
A tank that will actually give you a day's reserve is really on the large side IMHO.
You'd only realise it had all gone wrong when the day tank is empty.
Don't forget that a tank can generate its own contamination, it only takes a little water from the air via the vent.

What system is ideal depends a lot on how you intend to use the boat.

I've found having a reserve in cans very valuable, it trades well.

 

[pmagowan]

Many people use the phrase 'day tank' to mean a fairly small tank.
A tank that will actually give you a day's reserve is really on the large side IMHO.
You'd only realise it had all gone wrong when the day tank is empty.
Don't forget that a tank can generate its own contamination, it only takes a little water from the air via the vent.

What system is ideal depends a lot on how you intend to use the boat.

I've found having a reserve in cans very valuable, it trades well.

When I talk about a day tank I mean the amount you would expect to use in an average day on a sailing boat. It does not need to be enough to go at cruising speed for 24 hours under auxiliary power but it could be quite a bit as this would be for a blue water cruising boat with a very varied use. You would realise it had gone wrong when the fuel gauge was low,exactly the same as you would in any system. In fact you would realise sooner as it would alarm to tell you that the system had failed and dropped down a level. The difference would be that with only minor intervention you could get it all going again without lugging jerry cans about. The vents would have desiccant in them so the chance of significant water contamination in a GRP tank would be low, certainly no more than for a normal system where the fuel is not pre filtered twice.

The problem is it sounds complex but it is really a simple system. All it does is automates some simple functions that humans are not very good at. It will always remember to check the fuel levels and then it is simply a matter of turning on a pump and opening a few valves. Since it is fail safe the worst case scenario is that you fail all the way back to a simple day tank setup. For instance, imaging the day tank is 80L. You start the day with a full 80L but you are having a very unlucky day. The Pi fails and you don't notice, the arduino carries on keeping the day tank topped up regardless but a bolt of lightning nearby causes this to fail so now you have to flick a switch to activate the pump but the battery explodes so now you use the manual pump but a gas fire melts it so now you simply have an 80L tank of fuel running under gravity.

of course, in all likelihood none of these things will happen so you will be able to look at your iPhone and see that the day tank is full and that the hold tanks have 200L spare capacity in them, the engine is at full operating temperature, the batteries are cool and charged and that 'with or without you' is playing on the stereo.

 

[prv]

you will be able to look at your iPhone and see that the day tank is full and that the hold tanks have 200L spare capacity in them, the engine is at full operating temperature, the batteries are cool and charged and that 'with or without you' is playing on the stereo.

...and the boat is still up on bricks with no interior because you're still building the overcomplicated engineroom systems

Pete

 

[pmagowan]

...and the boat is still up on bricks with no interior because you're still building the overcomplicated engineroom systems

Pete

No, the boat isn't even designed yet. I like to have multiple projects on the go so am happy to learn programming and electronics while I am rewiring my house, building my garage, porch, orangery and designing my boat. I need to make my windows first, then design and build my boat shed. Trust me, the electronics are not complicated in comparison to all that lot and will likely to be finished well before the boat. I am also planing on building a hobbit house in the garden with a microbrewery which will require the same electronics for a 3 tank mashing and brewing system. I already have a pile if LEDs flashing away to my programming so how hard can it be

Ps I am going to redeck my current boat in the middle of all that.

 

[GHA]

A voice of enthusiasm amongst the doomsayers.... (from a blue water..)

I'd go for it, can't get my head round the 2 keel tanks, maybe just some sort of buffer valve so they just slosh and keep roughly level?
I've gone the day tank route on a recent refit, wasn't that hard but now benefits are...
Gravity fed engine, can polish fuel or even if water does get into the day tank dump it through a filter back in the main and then back through a filter into the day tank. Also, fill either tank for jerry cans or fill a jerry can to help out your new best buddy boat who is low 100 miles from landfall.
Haven't sorted out control of the pump yet or display from the gauge, but you've given me an idea, arduino!

I'm a big fan of arduinos, there's one controlling the fridge at the moment, (hopefully, not been near the boat for a while), very cheap, quite easy to program and bluetooth modules cost little as well so the option of tablet control/display is easy to implement. For day tank pump I'll have a switch and egg timer handy as well.

For more general boat hub arduino's might be worth looking at instead of a Pi, they can draw next to nothing and I/O seems easier (for a fairly novice like me).
Data logging is easy with an sd card, log your battery voltage, fridge temp, exhaust temp, engine running hours, fridge on off hours, so much just with few wires here and there.
Might be worth checking out the arduino yun, not long out, an arduino Leonardo and Linux system in one box, log the lot with just the arduino chip then boot Linux to upload to Dropbox when wifi is available.

For the naysayers, any blue water boat is loaded with silicon chips already, a few more won't make much difference so long as the basic systems can be run manually, nice enclosure and a good voltage regulator, no different from the chart plotter, gps, laptop, etc etc...

Also, for navigation google cubietruck, I've had it running OpenCpn nicely

As for the labour, most of it probably is figuring out how to make it all work, which is fun so obviously a plus

 

[charles_reed]

No, the boat isn't even designed yet. I like to have multiple projects on the go so am happy to learn programming and electronics while I am rewiring my house, building my garage, porch, orangery and designing my boat. I need to make my windows first, then design and build my boat shed. Trust me, the electronics are not complicated in comparison to all that lot and will likely to be finished well before the boat. I am also planing on building a hobbit house in the garden with a microbrewery which will require the same electronics for a 3 tank mashing and brewing system. I already have a pile if LEDs flashing away to my programming so how hard can it be

Ps I am going to redeck my current boat in the middle of all that.

Have a look at a few designers' notes - Marchaj (x2), Robert Perry, Ted Brewer, Norman Skene. More authoritative than us poor users and will keep you gainfully occupied for some time.
I'm assuming you can't read technical French, because there are some very good design books in that language.

 

[pmagowan]

A voice of enthusiasm amongst the doomsayers.... (from a blue water..)

I'd go for it, can't get my head round the 2 keel tanks, maybe just some sort of buffer valve so they just slosh and keep roughly level?
I've gone the day tank route on a recent refit, wasn't that hard but now benefits are...
Gravity fed engine, can polish fuel or even if water does get into the day tank dump it through a filter back in the main and then back through a filter into the day tank. Also, fill either tank for jerry cans or fill a jerry can to help out your new best buddy boat who is low 100 miles from landfall.
Haven't sorted out control of the pump yet or display from the gauge, but you've given me an idea, arduino!

I'm a big fan of arduinos, there's one controlling the fridge at the moment, (hopefully, not been near the boat for a while), very cheap, quite easy to program and bluetooth modules cost little as well so the option of tablet control/display is easy to implement. For day tank pump I'll have a switch and egg timer handy as well.

For more general boat hub arduino's might be worth looking at instead of a Pi, they can draw next to nothing and I/O seems easier (for a fairly novice like me).
Data logging is easy with an sd card, log your battery voltage, fridge temp, exhaust temp, engine running hours, fridge on off hours, so much just with few wires here and there.
Might be worth checking out the arduino yun, not long out, an arduino Leonardo and Linux system in one box, log the lot with just the arduino chip then boot Linux to upload to Dropbox when wifi is available.

For the naysayers, any blue water boat is loaded with silicon chips already, a few more won't make much difference so long as the basic systems can be run manually, nice enclosure and a good voltage regulator, no different from the chart plotter, gps, laptop, etc etc...

Also, for navigation google cubietruck, I've had it running OpenCpn nicely

As for the labour, most of it probably is figuring out how to make it all work, which is fun so obviously a plus

Glad to see there are some people out there that are into these things. I was also thinking of having the fridge kept an eye on. I can see loads of tasks that the arduino could be good at but I think it needs to be a Pi sat the higher level as the arduino is not really good at processing for a good user interface. I like the idea of a central unit that manages everything and stores all the information for diagnostics but also lets you play your music.

 

[pmagowan]

Have a look at a few designers' notes - Marchaj (x2), Robert Perry, Ted Brewer, Norman Skene. More authoritative than us poor users and will keep you gainfully occupied for some time.
I'm assuming you can't read technical French, because there are some very good design books in that language.

I have been reading loads of design books including Brewers etc. This is one of the things I like about DIY is that you expand your knowledge of everything.

 

[halcyon]

To do this I will have to split them into two tanks, one on either side running along the hull. They will be integrated into the hull for most efficient use of space and also because they then provide a 'double-hull' for added protection from sinking. I will then need a 'day tank' for feeding the engine. This could be considerably smaller and above the engine for ease of bleeding and to prevent air sucking into the fuel lines.

All this will need to be controlled throught the transfer of fuel from tank to tank. I have just discovered microprocessors and controllers such as the raspberry pi and the arduino and was thinking that this could be useful. My idea is to have a fuel level sensor in each tank, an 'in' valve (motorised) and an 'out' valve for each tank and a pump with a racor filter on it. The tanks will be connected in parallel with the pump so that transfer of fuel from one tank to another simply involves opening the 'out' valve for the tank giving fuel and opening the 'in' valve for the tank recieving fuel and then turning the pump on.

Is there a reason. other than the challenge, for using a micro ? as the above can be done with 2/3 LM324's and a few transistors.

Brian

 

[Deleted member 36384]

The Rival 41C is a centre cockpit and she has two large diesel tanks bonded into the hull on each side of the centre cockpit. The port tank forms a shelf in the cockpit locker and the starboard tank forms the base for the pilot berth in passageway to the aft cabin. The helm's seat runs across the back of the centre cockpit of course and I have thought that this would make a good day tank. The tanks do not interfere with the other areas of the yacht. My water tanks are in the encapsulated keel.

I think I would want to keep this simple and just use a manual pump similar to a Whale Gusher Urchin. It would not take many strokes to fill the day tank or maintain a good level. The lightening strike that wipes out your electronics could effectively make your fuel transfer system redundant - yes I know and the starter motor.

 

[pmagowan]

Is there a reason. other than the challenge, for using a micro ? as the above can be done with 2/3 LM324's and a few transistors.

Brian

No, the arduino is just a simple way in to electronics and programming. Once I have it completely designed on the arduino it is probably possible to shrink it to a small chip or two.

 

[pmagowan]

The Rival 41C is a centre cockpit and she has two large diesel tanks bonded into the hull on each side of the centre cockpit. The port tank forms a shelf in the cockpit locker and the starboard tank forms the base for the pilot berth in passageway to the aft cabin. The helm's seat runs across the back of the centre cockpit of course and I have thought that this would make a good day tank. The tanks do not interfere with the other areas of the yacht. My water tanks are in the encapsulated keel.

I think I would want to keep this simple and just use a manual pump similar to a Whale Gusher Urchin. It would not take many strokes to fill the day tank or maintain a good level. The lightening strike that wipes out your electronics could effectively make your fuel transfer system redundant - yes I know and the starter motor.

But surely this is the same reasoning for not having a chart plotter. The electronics could fail under extreme circumstances but they do not prevent you using the fallback method such as paper charts or a hand pump. They do, however, make life easy 99% of the time.

 

[sarabande]

sorry, PMG, but just because it CAN be done, does not entail that electronics is the best way to do it. If long-distance and autonomous cruising are likely to form part of the future, then bullet-proof systems are a must.

I concur with the day tank/ main tanks principle, and would really like to see a small fuel polishing system incorporated so that - perhaps manually - you can de-crud a grubby fuel supply.


My main sense of discomfort stems from reliance on electrons and clever programming.

What do you really need to measure ?

1 where the fuel is (day or main tanks) and

2 how much is in each tank


What do you really need to do ?

1 insert fuel into the main tank of your choice

2 transfer fuel from main to day

3.1 switch to an alternative clean day tank in case of contaminated fuel or

3.2 remove dirty fuel and insert clean into the day tank

4 get access to all the tanks to clean them


Simple, robust, mechanical float sensors or fuel sight gauges which do not, and never will fail because you have run out of electricity; a piping system which enable you to pump manually from main to day (are you ever going to use more than 50l in a day ?), with valves to isolate a dodgy tank; piping from the return from the engine to the tank of your choice; the ability to transfer from one main tank to the other (say on a long long trans-ocean tack); a 'last-chance' source of impeccably clean fuel held in reserve; and finally fuel filter bodies that are very easy to access and change, with very common filters.


I understand and admire why you want to make a clever system, and it would be very satisfying to do so, but for a mission-critical system, eliminating potential points of failure is the priority.

Can you really guarantee the MTBF of your home-designed and -made system ?

 

[fisherman]

KISS! I approve the day tank system, but ones I have experienced were designed for someone to go down into the engine room to pump fuel up to the day tank several times a day, at the same time checking out everything else. While not advocating this as ideal I would say you are over complicating things. Simple feed up/overflow back, constant polishing, gravity feed, option for emergency top up, but please, no complex electrics.