The great MVHR duct debate.

[MikeSharp01]

The architect has now worked out most of the MVHR duct routes around our design and so we have to decide which ducting method to use. We could use the Metal ducting for the spine runs but then we have to go radial from the spine ends, and in one case middle. Having have a variety of discussions and looking at it I am unsure about which radial duct system to use. I guess it comes down to semi rigid (as Renson for instance) or perhaps the fully flexible stuff (such as BPC). There is also the choice of plenum / diffuser to ensure we have access, adjustment, match the duct cross sectional areas and meet the aesthetic requirements. There are several assumptions and issues that I can see with this choice:

 

1. Assumption: that the supply and extract run together so the extract removes the same volume of air as the supply provides (appreciate there will be some small gains / losses)  so no positive/negative pressure situation.
2. Assumption: part F purge rule applies to the internal 'wet rooms' (IE no windows) and we have extended this so that we can purge all the wet areas according to part F all calculations here are based on boost mode to part F (2010).
3. Assumption: that we provide relatively open routes between supply and extract, door gaps for instance - so noise is limited and differential pressure is minimised (assumed zero plenum to plenum for calcs).
4. Assumption: If the MVHR units extract line were perfectly (as in perfect collision) connected to the supply line there would be no back pressure. (IE internal resistance is zero)
5. Assumption: Vent Axia say a well designed system, for a four bedroomed house will have a back pressure of 160 pascals so assume 180 pascals for our system as perfect we are not.
6. Main Issue: Noise, we want to minimise noise from the system so the air velocity wants to be as low as possible for the required flow rate. Our volumes are not huge even in purge mode I am modelling on 75 l/s in boost mode (270m³ / hr.) and we are working on a 400m3 / hr unit so in boost mode running at about 75% capacity - without back pressure adjustment see below. Worst case (model) is 2.11 m/s air velocity in extract pipe (assume equivalent of 100mm internal dia duct) from the main bathroom.
7. Issue: Simplicity of install, we want a simple install that we can clean out, as and when (hamsters etc),and that can get into small spaces so we can run up it stud walls etc (hence Renson or BPC type ducting)
8. Issue: should I include a calculation that uses the diameter, surface area, and perhaps roughness of the internal surface to give me some approximation of the back pressure I will get from a length of the relevant duct and thus the increased noise, as a result of more work, from the fan in the MVHR unit as most of them provide a performance curve against back pressure. Rather than assume the vent axia value?
9. Issue: Cost benefit - I guess I can decide on that!

 

I guess what I want to understand / be clear about - and having made a few whistles in the past, albeit steam, is what the effects of the duct choice will be on the noise and how the plenum / diffuser combinations 'tune up' against the two duct types.

 

 

Ventilation calcs Part F.pdf

Edited February 5, 2017 by MikeSharp01
Sorry forgot to add my spreadsheet model

[Jeremy Harris]

All I can comment on is the way I designed our system, based on first principles of air flow in ducts, the building regs requirements (which are more than a little odd in places) and our personal desire for a silent system in normal use.  I also wanted to DIY the install, and had no skills in working with ducts.

 

I started with putting terminals on the CAD plans where there would be the longest possible air flow route through a room, to allow the longest transit time and ensure the best possible chance that the air coming into the room would diffuse out across the whole room volume.  The logic of this applies equally to extract terminals, as they are pulling air in from under a door, or through a doorway.  For example, in our living room the entrance door is in one corner, so the air feed terminal is in the ceiling of the diametrically opposite corner.

 

Next I looked at duct systems, and concluded that the Swiss designed HB+ (made under licence now by around three or four companies) 75mm OD, "semi-rigid" smooth bore plastic ducting was the easiest to DIY install.  I bought a sample terminal and duct fitting on line, had a good look at it, and decided that's what I'd use.  Running the ducting was, realistically, something that could only be outline designed first, as the alignment of the ribs in the Posijoists dictated where the ducts had to go, and my CAD plans didn't have the exact position of every joist rib!  So, I put all the ducts in, and for safety ran double ducts to the kitchen extract terminal (the HB+ original design includes provision for two ducts per terminal - not all the copies of it do).

 

Once all the ducts were in and I knew the exact fitted length, the radius and position of curves in the ducts and the exact positions of the manifolds, I went back to first principles to calculate the flow rate in each duct at the building regs mandated ventilation rate.  I found that I was generally within the low noise range of under about 2.5m/s in most of the ducts, except for the longest run, to our bedroom, where I was around 2.8 m/s at 3/4 of maximum boost, and over 3m/s at full boost.  I decided to just see how it panned out, as I could, if need be, run another duct to that bedroom along the eaves space if it turned out to be too noisy.

 

I hooked up the MVHR, mounting it on old-style Mini exhaust rubber bobbins (there's a bit here about it: http://www.mayfly.eu/2014/04/part-thirty-mvhr-details/)   and made an error, in not fitting the recommended silencers in the main ducts that feed the manifolds.  This was me thinking the plenum effect in the manifolds would act as silencers, I was wrong!  I then set about setting the system up, by fitting restrictor plates inside the manifolds to control the flow rate in each duct.  I made up a home made flow meter, by buying a (still in calibration) second hand hot wire flow meter from ebay for around £30, and then gluing some PVC duct reducers and pipe together to make a cone and flow straightening pipe to take the flow meter.  Here's a photo of it in use -  anyone is welcome to borrow it, but it's no longer officially in calibration (not that I think for one moment that matters much - these things are very stable unless damaged, and BC seem not to be bothered):

 

 

To be honest, balancing the individual ducts was the worst part of the job, as it meant doing a measurement at a terminal, seeing what the flow was, running upstairs, opening the manifold, adjusting the restrictor plates by punching different rings out, then going back and repeating the exercise.  What's more, there are interdependencies, because the plenum pressure changes with every flow adjustment................  Thankfully it's a one-off task, and because there's no flow adjustment at the terminals, the noise is lower and you can pop them off and stick them in the dishwasher to clean them, without fear of messing up the flow rate settings.

 

I mentioned earlier that I didn't fit silencers, even though Genvex said they should be fitted to the room side of the unit (fresh air feeds and extracts).  The reason was there wasn't room to fit ready made silencers, and as above I wrongly thought the plenums would do the job.  So, I ended up making custom built absorption silencers, lined with acoustic foam (the fire resistant stuff).  They aren't pretty, but they work incredibly well, there is no noise at all from any duct, except when the system is on full boost, and even then the noise is only just audible, and not enough to disturb sleep (not that it's likely that anyone in our household would be sleeping when the things on full boost, anyway.  Here are some photos of one of my custom built silencers, the extract side one.  The fresh air feed one is just a rectangular box, lined with acoustic foam and sat directly onto the top of the extract manifold, which is on the floor, out of sight behind and to the left of the MVHR.

 

Front view of extract silencer

 

Side view of extract silencer

 

Internal view with the front cover unscrewed

 

Not sure if this helps or hinders, Mike, but we did end up with a system that changes the air in the house about once every 2 1/4 hours, seems to keep it very well ventilated and is pretty much silent.  Several people here have visited and heard it running, and they may well be able to give an independent view as to whether they thought it was as quiet as I think it is!

 

 

 

 

[MikeSharp01]

Cripes Jeremy I only just posted this and an amazing response. I had read your stuff on ebuild many moons ago and was looking for SFP curves for your Genvex unit just now. It seems that SFP is one of the keys to quiet but I cannot calculate / estimate how much although I understand why it's important. Skuttlebut says below 2 is ok but below one would be better. I will digest your response , cooking supper right now, and tweak my plans.

Edited February 5, 2017 by MikeSharp01

[Jeremy Harris]

Frankly I'd not get too hung up on duct flow rates if you go for a radial system like ours.  You can calculate things out to the nth degree, but there are so many variables around in practice that at best your calcs will be within around 50% of the measured flow rates.  I spent an awful lot of time doing calculations, only to find that in practice they were miles out!

 

A good example is the kitchen extract.  To meet the max extract needed by BRs, I fitted two ducts to that terminal.  It's running in balance with one of those ducts throttle down to the smallest restrictor disk and the other with a disk that's around 2/3rds the maximum, so a single duct would have easily met the regs.

 

One thing I forgot above - don't make the mistake I made of fitting the intake on one external corner of the house and the exhaust on the other.  In our case the slightest breeze throws the balance and flow rates out a lot, and I've got some cowled stainless external terminals that I intend fitting, with the intake cowl pointed away from the wind that blows along that wall, to try and fix the issue.  Practically it doesn't seem to have any noticeable effect, but it does make the flow rates at the terminals leap around a fair bit when you are measuring them.

[MikeSharp01]

Thanks again, mid supper prep, I guess I want to show the BR people that I have thought things through and in the end if I get the full 4 ACH in boot mode, for the internal WC/ en-suite on about the 75% speed setting and under 2 SFP I will be happy. Good point on the location of the inlet / outlet of the unit. My plan was run then straight up, cowell them both and have the outlet above the inlet by a meter or two. I am not sure I can get them separated much in X or Y given restrictions of the plot boundary.

 

 

Edited February 5, 2017 by MikeSharp01

[Jeremy Harris]

This may help, it's the report I compiled for BC, to show that the ventilation system complied with the regs.  It's complete over-kill, building control would have been happy with far less detail!

 

Mechanical Ventilation and Heat Recovery system Test Report.pdf

[MikeSharp01]

Thanks Jeremy. That looks like feasible document for me to produce. 

[joe90]

I too have saved it to input my own data and present to the BCO. Planning my own MVHR is next and like Jeremy want to DIY as much as possible.

[DH202020]

6 hours ago, JSHarris said:

This may help, it's the report I compiled for BC, to show that the ventilation system complied with the regs.  It's complete over-kill, building control would have been happy with far less detail!

 

Mechanical Ventilation and Heat Recovery system Test Report.pdf

I need to balance my system, I've moved in now (still to build garage) and MVHR  works ok but not technically balanced.

 

My vent axia system is quite at normal running speeds, even on boost it's not that loud when extracting through the plenums.

 

I DIYed the flexible ducting using the plan the company designed for me, no rattles or vibrating, house is very comfortable, towls dry overnight, slightly cooler upstairs, which is nice in the bedrooms (about 1°). Only heating is underfloor on the ground floor.

 

Jeremy: I'm not sure how you work your gadget, but if you're willing to lend, I will give it ago using your PDF template.

This is not my forte but, I guess, I've built a detached house...how hard can it be??

 

If any one has a step by step (fool proof guide for idiots) be very interested

 

Thanks David.

 

[Jeremy Harris]

Sure you can borrow it David.  It's pretty easy to use, but the readings do move around a bit as the conditions change outside, so ideally choose a still day to do it.  It's got a bit of felt glued around the top cone to press up against the ceiling without causing damage.  IIRC, the top cone is a 150mm PVC duct adapter, which fitted over our terminals OK, and I think should fit over most domestic terminals.

 

If you PM me your address I'll chuck some new batteries in it and pop it in the post to you.

 

The process I used was to go around and measure all the flow rates, fresh air in and extract at full boost, and write down the average readings I got.  I then looked at how these compared with the regs for extract rates, and guessed at some values to reduce each by, as they were all going to be way over the regs requirement - it was really about getting the ratios between extract rooms about right at this stage.

 

I then adjusted the extract duct flow rates to get them in about the right extract ratio, and didn't worry about the fresh air in ones.

 

Next, I turned the system down to normal speed, and measured again, adjusting the fresh air in feeds to get the sum of the extracts = sum of the fresh air feeds.  This meant the system was in balance, in that the air in was the same as the air out.

 

Finally I checked the extract rates again at full boost to be sure they exceeded regs requirement (I knew they would, as they were close to it at trickle ventilation rate).  I also checked the sum of the extract rates with the sum of the fresh air in flow rates to see if the system was still in balance at full boost.  Mine was out by around 2% at boost, IIRC, which I decided was good enough.

Edited February 6, 2017 by JSHarris

[joe90]

With regard the MVHR duct to outside air, is there a special duct for this purpose. I have to get my duct in whilst building the roof/wall junction to make sure it's airtight. I have found a supplier of aluminium flexible ducts of various sizes near where I live and don't want to install something that's not fit for purpose. Also the vents will be fixed to the horizontal barge board, is there a particular type of terminal that is best for this?.

[SteamyTea]

How easy would it be to make a 'hot wire' mass flow meter to work with a RPI.

They are only a warmed thermistor, a Wheatstone Bridge and a temperature sensor.

So there would need to be an AD in there somewhere (or use an RC circuit for timing).

 

Just thinking if they could be made cheap enough, one could be fitted in each duct and then adjust till all the numbers are right.  Got to be better than running around the house all the time (and up in the loft).

[Jeremy Harris]

48 minutes ago, joe90 said:

With regard the MVHR duct to outside air, is there a special duct for this purpose. I have to get my duct in whilst building the roof/wall junction to make sure it's airtight. I have found a supplier of aluminium flexible ducts of various sizes near where I live and don't want to install something that's not fit for purpose. Also the vents will be fixed to the horizontal barge board, is there a particular type of terminal that is best for this?.

 

I used short sections of 150mm PVC duct through the external walls, bonded in and sealed, then used brown plastic external grills that just fit on to the 150mm duct and screw to the wall.  I chose brown as mine are on the larch cladding, plus I needed to make up some wedges to deal with the slope of the cladding overlap.  I'm about to replace them with some cowled stainless terminals, for two reasons.  One is the imbalance problem on windy days I mentioned earlier, the second is that the intake grill is hard to clean on the plastic terminals - it collects fluffy seeds and spider webs. 

 

These are the plastic grilles I have now: http://cart.vacuumsdirect.co.uk/index.php?p=product&id=377&parent=46

 

and the stainless ones I've bought are 150mm versions of these (but I bought mine on ebay for a fair bit less!): http://cart.vacuumsdirect.co.uk/index.php?p=product&id=392&parent=46

 

Internally I used the fairly rigid aluminium flexible duct to connect to the MVHR, not the very flexible stuff with the wire spring inside, as that has a fair bit more flow resistance.  This is the flexible stuff I used: http://cart.vacuumsdirect.co.uk/index.php?p=catalog&parent=38&pg=1

Edited February 6, 2017 by JSHarris

[Jeremy Harris]

37 minutes ago, SteamyTea said:

How easy would it be to make a 'hot wire' mass flow meter to work with a RPI.

They are only a warmed thermistor, a Wheatstone Bridge and a temperature sensor.

So there would need to be an AD in there somewhere (or use an RC circuit for timing).

 

Just thinking if they could be made cheap enough, one could be fitted in each duct and then adjust till all the numbers are right.  Got to be better than running around the house all the time (and up in the loft).

 

Not as easy as it seems!  I have a birectional micro-thermistor flow sensor, retrieved from an old glider variometer.  It has two very, very tiny thermistors, barely visible to the unaided eye, mounted in a hole so that one gets cooled by air flowing one way and the other gets cooled by air flowing the other way.  They are self-heated by a constant current power supply, that has to be accurate and ideally temperature compensated as well.  I made up a circuit using an instrumentation amplifier (an INA122) to amplify the tiny voltage from this half-bridge and feed it to an A to D on a microcontroller, and added temperature compensation separately, using a DS18B20 sensor, as I just used an off-the-shelf current source, an LM334, and that isn't temperature compensated.  The voltage change is non-linear with air flow velocity, so needs to be corrected, and that means you need a calibrated air flow meter to set the thing up.  In my case, I had access at that time to an aircraft instrument test set, which includes a variable speed pump and calibrated flow and pressure sensors, so I was able to just connect the two tubes and do the calibration that way. 

 

It would certainly be possible to make something like a mini-wind tunnel, with a bit of duct, some "egg box" flow straighteners and a model aircraft brushless motor, prop and speed controller.  This should be able to create a linear airflow at the right sort of air velocity range, and then something like the flow meter I've already got could be used to calibrate the motor speed against flow velocity.

 

All told it would be a bit of work, but is certainly do-able.  The RPi A to D should be OK, and the RPi can read one-wire devices like the DS18B20 as well, so all you need is a bit of analogue front end stuff to condition the signal to a level that the RPi A to D can work with.

[SteamyTea]

May have to set up a thread in Koffins Korner about this. I just get the feeling that there should be an easy and cheap way to monitor airflow.

 

 

[joe90]

Thanks Jeremy, I found these on EBay http://www.ebay.co.uk/itm/Circular-Stainless-Steel-Air-Vent-Grille-Covers-High-Quality-Ventilation-Grilles-/131165426150?var=430378594662&hash=item1e8a1191e6:m:m7BE1CJNES6LS-xIMFGBbMQ as I don't need bullnose, what do you think?

[martinwinlow]

Wow.  My input having done a 5 bed MVHR DIY a few years back...

 

Firstly, I can't help but think most people *grossly* over-engineer the design of these things.  You *can* spend hours and hours designing the b-jesus out of a MVHR system - 'meeting the regs', sound-proofing etc, etc... but, in reality, ultimately you will not be changing the air at anywhere near what the regs say you 'should' and duct noise really is not going to be an issue at the air speeds involved (assuming you avoid rigid ducts - see late and even balancing is more art than science). 

 

For an average 3/4 bed house, just stick to 4" ducting between the heat-exchanger box and vents (and use insulated flexible ducting as it is very cheap and easy to use and you won't get condensation issues that you may with rigid, uninsulated ducting).  Use round, plastic ceiling vents/diffusers  with adjustable flow rates which are dead simple to install, clean and balance (if you feel the need).

 

NOTE:  Get a ventilation unit with summer by-pass facility - I can't believe these things are sold without this facility as otherwise, in summer, you either have to have all the windows open all the time or the house becomes unbearably hot due to the heat recovery thing still going on.  You also would have to deal with bathroom/kitchen extracting being operational without any way to stop the house warming up, again, without opening the windows.

 

Bear in mind I was doing a new-build with a central services area and so all my ducting runs were kept to a minimum length and were easy to integrate.  If you are retrofitting then rigid ducts may make much more sense.