Flat roof make up

[Tony L]

@Gus Potter& dealing with your previous post:

 

I am grateful to you for taking the time to explain your thoughts on this - including those I don't like & those I disagree with.

 

I don’t have time to explain the history of how I got to this point, because I have big demands on my time at the moment, in addition to the self build project, but just taking a few of your points in order, from above.

 

The SE has been impressive, so far.  He seems completely undaunted by this project.

 

My architectural technician (& he’s not the first one I’ve had on this project) is a huge disappointment.  If I start to explain why, I’ll be here all night. 

 

Agree: his designed insulation envelope is not good enough.  I am having to work to fix this – not something I expected to have to do.

 

Rain will fall straight into the cavity behind my cladding (will be composite or Millboard type, so the top of the cladding is OK getting wet)  – some advice says a 6mm gap above the ventilation void (which will be 50-100mm, depending on how much building out we do to get the cladding to where we want it to be, will be enough to vent & will not let much rain in.  I could easily have the drip edge from the flat roof running over the top of the cladding, whilst allowing 50mm vented space all the way out of the top of the vent cavity then under the drip edge.  It’s not the look I wanted, but if it’s necessary, that’s what I’ll put on the drawings.

 

I’ve got draft drawings from the SE, so I know where all the steels go & this is how I know my rafter depths are different for the two pitched roof sections.

 

Re: “can I get someone to build & at what cost”.  I don’t see how I can stop & do that now.  I need to get the drawings done properly then I can get a BoQ/build cost.

 

“blindly” I don’t agree with this characterisation of my approach to this project,  although I don't mind you telling me this is what you think.

 

“If I go back to the beginning, how much will that cost me?” – difficult to say in financial terms & I'm not going to waste time thinking about this right now, but “going back to the beginning” would cost me the relationship I have with the person who’s paying for the other half of all this.  Due to a looming planning permission deadline, the old house was demolished 2024.  Self building was her idea, fortunately.

 

 

 

[Tony L]

2 hours ago, Gus Potter said:

internal clearance for stairs, ventilation at window heads and maybe fire escape /

Stairs have more headroom than building regs require - so no problem bringing the roof at the front back towards the centre of the house.

 

Ventilation - I'm aiming for good airtightness & MVHR - you have possibly been misled by the MEVs the AT drew in, despite my clear brief, "air tight, + MVHR".

 

Fire: I read the building regs document.  I think my design is compliant.  My only concern is BC may demand doors across my kitchen, that opens up into the hallway & stairs.  No big problem if the doors have to go in here - it will just be an unwelcome expense.  Also, most of the steels will have to have intumescent paint, rather than be boxed in, because they have joists attached.

[Gus Potter]

34 minutes ago, Tony L said:

Thanks

Hi Tony. 

 

I can see you are putting in a lot of effort into this. I appreciate that your teck may have not "captured" the complexity, but how much are you paying? You get what you pay for to some extent.  You sound like a hands on person so if you have just paid you technician from the neck down then what do you expect? 

34 minutes ago, Tony L said:

I thought he'd know how to do all this stuff. 

This is a hard thing to learn, it takes years. To provide a bit of context. I was a building Contractor for 20 years, went to uni at the age of 40 to become an SE did training and now I've progressed to the point where I'm confident / competent in both SE, Architectural design, whats buildable and cost aware. That is a 40 year journey and every day is a schoolday for me. BH is a great resource. There are folk like you proposing this kind of innovative design. Your sketchup model looks brillaint!

 

20 minutes ago, Tony L said:

The SE has been impressive, so far.  He seems completely undaunted by this project.

 

It's not that hard if you go about it in the right way. Your SE sounds experienced, pragmatic, some SE's are actually sympathetic! We are not all horrible, sometimes we dish out tough love when it's clear that is what is required to give the Client a bit of a jolt and protect their interests. 

20 minutes ago, Tony L said:

“If I go back to the beginning, how much will that cost me?” – difficult to say in financial terms & I'm not going to waste time thinking about this right now, but “going back to the beginning” would cost me the relationship I have with the person who’s paying for the other half of all this.  Due to a looming planning permission deadline, the old house was demolished 2024.  Self building was her idea, fortunately.

So what to do? If you want to keep the rest of the team on board then in your own head go back to basics. Bluntly you can probably fix it by reconfiguring what is going on under the bonnet. The planners have no interest in the structure, insulation or what's going on inside the building. All thye are interested in is what it looks like from the outside. If you increase the occupancy that is material as it could impact on the parking provision, but all we are looking at here is seeking a buildable, cost effective design that is not going to fail due to say condensation, water ingress etc.

 

You asked about cold formed steel joists. You are right they can be a bit of wicked cold bridge, but on a warm roof you keep them on the warm side of the insulation. Dormer roofs tend not to be that big span wise. SE wise we need to check the bearing on the dormer face. But here we only need about 50 - 60 mm, this is driven by the fixing requirement into timber to get enough timber edge distance. The bearing strees does not govern usually on a msdest dormer roof.

 

You can get some insulation in to stop the ends of the joists getting cold. Also the cold formed joists may be only 1.2 -1.5mm thick. So while the rate of heat transfer in steel may be much more than a timber if you look at an exposed timber joist end which may be 45mm thick cf a 1.5 mm bit of steel exposed is not as bad as you think. Most folk think about big universal steel beams.. but here we are using thin steels that have a small expose cross section.. it's ecclectic but part of the art of design. 

30 minutes ago, Tony L said:

I don’t see how I can stop & do that now. 

I think you can and you need to find a diplomatic way of doing so. It's just a move from the concept to detailed design stage after all! Detailed desin always uncovers unexpected cost elements, design is an itterative process. I pointed out why I think you are going to suffer, all your are doing is developing the design and details as you learn about the best and most cost effective way of achieving what you want.

 

 

 

 

 

 

 

 

 

 

 

[Gus Potter]

15 minutes ago, Tony L said:

you have possibly been misled by the MEVs the AT drew in

Post overlapping here Tony.

 

I'm trying to warn you of the challenges you face. I have not considered your internal ventilation, only that the structure is going to work insulation and ventilation wise.

15 minutes ago, Tony L said:

Fire: I read the building regs document. 

The fire protection is always a challenge. Your house is only two stories and detached so pretty benign. 

 

In your case so long as any roof steels are not contributing to the stability of a floor or structural load bearing wall they may only need minimal protection. Talk to your SE and you may find it all works boxed in by ordinary plasterboard and buried in the roof dpeth.

 

 

[Tony L]

Thanks for all this, @Gus Potter.

 

10 hours ago, Gus Potter said:

You can get some insulation in to stop the ends of the joists getting cold. Also the cold formed joists may be only 1.2 -1.5mm thick. So while the rate of heat transfer in steel may be much more than a timber if you look at an exposed timber joist end which may be 45mm thick cf a 1.5 mm bit of steel exposed is not as bad as you think. Most folk think about big universal steel beams.. but here we are using thin steels that have a small expose cross section.. it's ecclectic but part of the art of design. 

 

I've designed all my metal web floor joists so they're inside the inner block leaf; those that meet the outer wall will be hung on ledger boards.  I've done this to make air tightness easier to achieve, & there's a thermal break benefit too.

 

With the ideas @Iceverge has provided, the tops of the rafters will all be insulated, but at the bottom they'll run from inside to outside over the wall plate, of course.

 

The SE has specified UB19 for the beam in the section picture you posted.  This is a fair bit chunkier than I was expecting - only because I can lift the mild steel box section back bone chassis of an old Lotus with ease, & they're very strong.  These UB19s (there are 6 of them) are in lengths of max. 4.55m, so that's 86KG to lift into place.   I think the top of a UB19 will be 7.9mm thick & the web will be 4.8mm, so it will be good to have insulation along the edges of these UBs, & it's worth considering what's going on at the beam ends too.  My plan was to wrap the beam ends that penetrate the inner cavity block leaf with VCL membrane, bring the membrane inside & fix it to the wall (I've got notes on how to affix). The cavity between the block inner & rendered block outer leaf is only 150mm & will be filled with EPS beads.  I was thinking it may be worth sticking a rectangle of super insulating something or other inside the cavity on the beam ends.  If anyone has any suggestions please let me know.  & is it worth sticking a block of high R-value something on the top flanges?  There's not much gable end wall going on top of these UB ends.  I looked for high R-value padstones, but I gave up this search & decided to rest the beam ends on regular padstones.

 

 

[Iceverge]

In general once stuff is inside the main insualtive layer then it's not worth worrying about extra minute thermal gains like insulating padstones or the like. 

 

As i can see you'll have 150mm EPS beads outside your UB so any gains would be very small. 

 

I would stop the inner leaf of the wall 100mm short of the external wall and allow the 100mm PIR to oversail the wall and meet the wall EPS beads. This will make a thermally very nice junction. 

 

In general I'm not a fan of too much steel work in domestic construction. It's hard to finish and fix to with typical builders tools. 

 

I would prefer precast concrete or on site poured lintels for the blockwork sections and gluelams for the roof. They're aesthetically quite nice in my opinion and leaving them on show may be cheaper and easier than building in a UB. 

 

 

In general the more I look at your house I can't but agree with @Gus Potter. There's a lot of obstacles to getting it to the standard you hope. It has the classic hallmarks of something that perhaps began more simple but had bits bolted on to satisfy a desire for something different or with a bit more flair. Then it sounds like planning appeared and imposed some limits that compromised the building. 

 

1 went through 3 sets of planning permissions before we build our house. Each one simpler in turn and it was still a pain as a first time builder to get anywhere close to the standards I wanted with efficiency etc.  Would you consider returning to the drawing board and designing out some of the trickier bits like reverse slope dormers and hidden guttering? 

 

 

 

 

[Iceverge]

 

Based in your design constraints a section like this would be much more buildable. 

 

No steps in the upstairs. 45 deg roof into a 3 deg section in the middle for sanding seam. likewise with the dormer. 400mm allowed for the walls and roof for plenty of insulation. 

 

 

[Tony L]

@Iceverge. Thank you again for your work on this.  I'm pleased you support the idea of "flat roofs" being standing seam at 3 degrees.  Let's hope the planners consider this a reasonable fall for my approved flat roof sections (I'm not intending to ask them).

 

I've been working on some detail drawings for the roof intersections this morning.  I have to go out now, so I'll aim to I'll finish these off & post them this evening, along with responses to the other points from your post above.

[Nickfromwales]

On 09/01/2026 at 12:17, Tony L said:

The cavity between the block inner & rendered block outer leaf is only 150mm & will be filled with EPS beads.

Can you not use slabs?

 

https://www.insulationuk.co.uk/products/150mm-superglass-superwall-32-cavity-wall-batt

[Tony L]

Thanks again for this, @Iceverge.  I like what you’ve drawn. Sorry, the roof can’t be at 45 degrees because this would stretch the flat roof section at the top & make the building look very different from the approved plans.

 

400mm EPS in the walls is appealing, but it would make the inside too small & it can’t happen because [I’ll be in big trouble with you when you read this] the substructure is already built.  For the benefit of anybody reading this whilst researching your possible self build, & to save @Gus Potter & @Iceverge the trouble, l will just explain: building the substructure before your superstructure drawings have been finalised is a very bad idea.

 

[Tony L]

On 09/01/2026 at 19:12, Iceverge said:

Would you consider returning to the drawing board and designing out some of the trickier bits like reverse slope dormers and hidden guttering? 

I’m sorry to make the task of resolving these problems more difficult but, there is no way I’m going back for a new planning permission.  I don’t want to spend time explaining the long history of how I eventually achieved PP for something that I actually wanted to build: I need to focus on what’s ahead.

 

On 09/01/2026 at 19:12, Iceverge said:

It has the classic hallmarks of something that perhaps began more simple but had bits bolted on to satisfy a desire for something different or with a bit more flair. Then it sounds like planning appeared and imposed some limits that compromised the building. 

That's not far off what happened, yes.

 

On 09/01/2026 at 19:12, Iceverge said:

I would prefer precast concrete or on site poured lintels for the blockwork sections and gluelams for the roof. They're aesthetically quite nice in my opinion and leaving them on show may be cheaper and easier than building in a UB. 

I don't know anything about site poured lintels.  I think I ought to minimise the number of technologies I must learn about to get this house built - I'd really want to learn about it because I wouldn't trust a builder to do it properly, if he thought I didn't understand the process.  My learnings about works up to DPC (mostly from BH) saved me at least £7,000 after the builder took advantage of BC's requests for changes & tried to rip me off.

 

I like glulams too.  Are these OK to stick through the inner block leaf, so they're flush with the cavity (wrapped in VCL, of course)?

 

On 09/01/2026 at 19:12, Iceverge said:

I would stop the inner leaf of the wall 100mm short of the external wall and allow the 100mm PIR to oversail the wall and meet the wall EPS beads. This will make a thermally very nice junction. 

Sorry, I can't visualise this.  Are you talking about the long inner walls that the rafter ends sit on or the gable end walls (that will have ladders going over them, for some roof overhang)?

 

[Tony L]

I’ve added to a drawing I found online to show how the ventilation & drip edges will work at the apex of a standing seam roof, at 3 degrees .

[Tony L]

& here’s a detail for where the 3 degrees centre section meets the 42 degrees section.  I’ve added in some L shaped plates (green) to prevent rain being blown up the roof & into the vent void.

 

This same detail could be used where the the 42 degrees sections meet the backs of the 3 degrees flat roof sections over the dormers.

 

 

 

[Tony L]

9 hours ago, Nickfromwales said:

On 09/01/2026 at 12:17, Tony L said:

The cavity between the block inner & rendered block outer leaf is only 150mm & will be filled with EPS beads.

Can you not use slabs?

 

Thanks for the suggestion, Nick.  Yes, I could do, but I like EPS beads because:

 

EPS will fit perfectly into all the awkward shaped edges, on top of cavity trays, into a badly built section of cavity that is 165mm wide (ref: one of my threads from months ago).

If my trench blocks get wet, I think wet/damp EPS will be better than wet/damp mineral wool.

It doesn’t have to be stored on site.

It won’t get wet during installation.

It won’t be installed by the brickie, who’s on a price for the job & who’s motivated to work quickly & cut corners (ie leave gaps in the insulation), so he earns himself a paid day off at the end of the job.

We’ll be able to start putting the roof on sooner.

Perhaps the brickie/brickies will enjoy laying the blocks & do a better job if they don’t have to faff around with fluff.

The internet says EPS will be cheaper (so I’m assuming, if it costs more it won’t be loads more).

The internet says EPS has a better R-value (& that will be as compared to the expertly & patiently prepared mineral wool that was lab tested to deduce its R-value).

I reckon I can procure EPS installation service myself, so I won’t have to rely on the builder to deal with it, nor pay him to watch it being done, nor pay his margin.

I get to choose who does the work & it will be easier for me to arrange to be there to oversee it when it happens 

@Iceverge likes EPS & I’m an @Iceverge superfan.