Where to start with this defect in our “preimum” home…

There has always been movement and noise from the flooring upstairs. The landing is bouncy, if you’re underneath in some rooms there’s loud cracking sounds whenever anyone walks across the floors upstairs, the staircase wobbles and the stairs themselves creak when you just walk past them nevermind when you go up or down. As with all the defects in our Bovis home we have had to fight tooth and nail to get to the bottom of the issues. In the first 6 months we had 3 separate visits with workmen pulling up carpet, underlay and boards to “fix” these issues. Here are just a couple of highlights of our experiences:

On either the second or third visit, contractors were fitting a herringbone configuration to stop the boards moving, part way through the day I went upstairs to see what was happening, I asked the contractors what they had found and one told me that they are not allowed to tell me – they had been placed under a gagging order! The work is going on in my home but I was not to be told what issues they had found. After this, we advised Bovis that if that ever happened again then the contractors will be asked to leave our property. Needless to say, this eroded any trust we had left for Bovis.

Another occasion, in a different room, contractors had conducted a similar exercise, after they replaced the underlay and carpet, I kept feeling a hard lump. Bovis advised that everything was fine. When the carpeting and underlay were once again pulled back for more “fixes” to be applied, an inspection tool was found between the carpet and underlay and miraculously the hard lump disappeared!

All bar one bedroom and the landing had multiple contractor visits, from grid screwing to the herringbone bracing – none of which resolved the issues. At this point, we were told by the then Area Customer Service Manager, that if the herringboning fixed the issue, they would return to do the other rooms, then Covid hit, after Covid, the same person told us that they had already done a load of work, the issues were in his opinion, within tolerance and they were not prepared to do any more. Clearly, we did not accept that answer and got the NHBC involved. Bovis then arranged for an independent inspection, a senior consultant from TRADA carried this out and wrote a detailed report. This report describes a floating floor in one bedroom, Caberdeck boards that don’t rest on the joists, joins that don’t end on the joists, joists set at different centres and levels, lack of glue, ineffectual noggins, ineffectual fixes and boards resting on a considerable amount of foam instead of the joists… but don’t take my word for it, here are some excerpts from the TRADA report:

The first inspection consisted of an initial assessment of the upper floor in order to establish
the levels of noise generation occurring in each room. This would be followed by intrusive
examinations carried out at selected locations on the upper floor in order to examine the
construction of the floor at each.

General

An overall inspection of the upper floor was carried out in order to determine the extent and
severity of the apparent noise generation experienced in each room under normal foot traffic.
Noise generation in the form of underfoot creaking at the time of the inspection was found to
be most noticeable in Bedrooms 1 and 3 and less so in Bedroom 5. In the case of Bedroom
4, relatively little noise generation was found when the floor was trafficked within the room,
however creaking became readily apparent downstairs in the Dining Room immediately
below Bedroom 4 whenever the floor in Bedroom 4 was walked on. Bedroom 2 was assessed
as exhibiting the least amount of noise generation.

In addition to the above issues of noise generation, the area immediately around the stairwell
and including the floor to Bedroom 4 adjacent to it, gave the impression of being noticeably
spongy when walked on. Attempts to measure any apparent downward deflection using a
laser level proved to be unsuccessful, however when the floor was walked over, the
horizontal line of the laser was seen to bounce noticeably relative to the line of the skirting,
suggesting that the floor around the stairwell behaved as a diaphragm and explained the
spongy sensation when the floor in this area was walked on.

Intrusive Examinations

A number of upper floor locations were selected in which intrusive examinations were
performed. The approximate positions of the different inspection points are shown in Figure
1 below:

The locations were selected in such a manner as to represent the different types of floor
reaction observed whilst ensuring that locations were distributed over a large area of the
upper floor in order to account for different alignments of the floor joists. At each selected
location a square of the Caberdek flooring measuring approximately 300mm x 400mm was
extracted between the joists to enable a direct visual examination of the floor construction to
be made.
The following examinations were made in the sequence of their execution:

Bedroom 4 (Location “A” Figure 1)

Examination of the surface of the floor showed that countersunk screws had been inserted
into the face of the Caberdek boards along the lines of the joists (Refer Plate 1). Extraction
of an inspection hole in the Caberdek board along the line of a joist showed the board was
in good contact with it and had been well bonded by means of a liberal application of adhesive
(Refer Plate 2).

The Caberdek flooring was confirmed as being 22mm in thickness machined all round with
an interlocking tongue and grooved profile.
Examination of the floor void showed evidence that the Caberdek panels were in good
contact with their supporting joists with frequent evidence of adhesive having been expressed
from between the joists and the boards. However installation of the Dining Room plasterboard
to the lower edge of the joists was considered to be inexpertly carried out with visible gaps
being present between the plasterboard and the joists (Refer Plates 3 & 4).
No noggins were found to be present between lines of adjacent joists apart from in the
immediate vicinity of the blockwork along the external wall and the line of header joists distal
to it (Refer Plates 3 & 4).

Landing (Location “B” Figure 1)

Extraction of an inspection hole on the landing area adjacent to the stairwell showed much
the same situation as that found in Bedroom 4, where evidence of good adhesive bonding of
the Caberdek panels was again found along the upper edges of supporting joists in addition
to their having been face-fixed with countersunk screws along the lines of the joist (Refer
Plate 5).
Again no intermediate noggins were seen to have been inserted between adjacent lines of
joists apart from those in the immediate vicinity of the exterior blockwork wall and the header
joist distal to it. Evidence of adhesive expression along the interface between the decking
and joists indicated good adhesion of the Caberdek with the joists although evidence of
somewhat erratic adhesive application was also found where, beads of adhesive seemingly
applied to the Caberdek panels, were offset relative to the line of the joists (Refer Plates 6 &
7).

Bedroom 1 (Location “C” Figure 1)

The Caberdek floor in Bedroom 1 showed evidence of earlier intrusive work having been
undertaken in the form of decking panel inserts having been fitted to make good areas of
removed decking. The Caberdek panels were also seen to be face fixed with countersunk
screws running at right angles to the line of the joists as indicated by construction drawings
for the property suggesting that intermediary noggins may have been introduced between
adjacent lines of joists. (Refer Plate 8).

Extraction of an inspection hole along a line of intermediate screw fixings coincident with a
header joint between adjacent decking panels showed the header joint to be unsupported by
a joist but with a noggin inserted beneath it, held in place by screwing through the panel
above it, as a measure to prevent uneven deflection occurring between adjacent decking
elements (Refer Plate 9).

Elsewhere, it was seen that header joints between Caberdek boards were only partially
supported by means of noggins fitted between the joists. These additional noggins were
found to be very loose fitting and fixed with a single nail driven into the upper flange of a joist
at each end. The nails were set back from the ends of the noggins and driven at an angle
into the joist flange meaning that their holding power was, in all likelihood, compromised (Refer Plate 10).

The noggins were found to be loose as a consequence allowing them to be
moved and dislodged by hand. Furthermore, where such noggins had been introduced, they
were seen not to be in uniform contact with the underside of the deck boards, meaning that
the decking was largely unsupported by them (Refer Plate 11).
Examination of the floor void in the opposing direction (i.e. toward the front of the property),
showed a similar scenario where intermediate noggins were found to provide incomplete
support to the deck boards as well as not affording complete lateral support between
successive lines of joists as evidenced by noticeable gaps between the noggins and joist
flanges at both ends of the noggin. Visual evidence was found suggesting that the top flanges
of I-joists did not always make uniform contact with the undersides of the Caberdek panels
as evidenced by the presence of gaps occurring between the decking and joist along the
flange line of the latter (Refer Plate 12).

Bedroom 2 (Locations “D”, “E” & “F” Figure 1)

Three inspection holes were cut into the floor in Bedroom 2. The first hole was cut in the
vicinity of the door leading into the room and showed the lower flange member of the I-joist
to be strengthened by means of a square sectioned timber reinforcement fixed to the
lowermost flange member which spanned the width of the room (Refer Plate 13).
The joists in the floor space of bedroom 2 were fitted with noggins in the vicinity of the external
blockwork wall and along the header joist distal to it. No other intermediary noggins were
found. The Caberdek boards were found to be well seated on the joists with good apparent
contact having been made between the decking and the upper flanges of the joists (Refer
Plate 14).

A second hole was cut in another part of the room to establish whether further reinforcement
measures had been installed within the same room given that Bedroom 2 exhibited the least
amount of noise generation.

The second inspection hole showed that no additional reinforcement elements had been
introduced to other joist members as far as was possible to determine. However removal of
the Caberdek decking along a joint line showed the joint line was not exactly coincident with
the line of the joist (Refer Plate 15). On closer examination the tongue and groove joint was
found to be loose fitting with no fixings or traces of adhesive in evidence (Refer Plates 16 &
17).
Inspection of the floor void revealed no significant differences in the construction method
used. However, further examples were seen where the Caberdek decking was not in direct
contact with the upper flange of the joist member but instead appeared to be supported by a
bead of foaming adhesive (Refer Plate 18).

A cursory examination of the stagger between header joints occurring between adjacent rows
of decking boards indicated that the header joints of the decking may not be adequately
supported given that the spacing between header joints was seen to exceed that of the Ijoists.
It was therefore decided to cut a third inspection hole in Bedroom 2 in order to verify
whether this was the case.

On extracting a third hole it was found that a double line of joists had been installed at this
location and even then the header joint of the deck boards was seen to be exactly coincident with the outer edge of the upper flange of the joists and not centrally supported. The header joint in this instance was found to be bonded with adhesive (Refer Plate 19).

Bedroom 3

Two inspection holes were cut in Bedroom 3. The first was cut in a location which had been
subjected to earlier remedial works as evidenced by the fact that the decking boards had
been grid screwed (Refer Plate 20). The second hole was cut in a location where no previous
work had been carried out.

The first inspection hole revealed the presence of a double joist one of which showed
evidence that the decking had been glued whereas the other showed no adhesive deposits
(Refer Plate 21).

Inspection of the void space via the first inspection hole on one side of the double joist
showed that no noggins had been inserted apart from one along the blockwork wall. The floor
decking also appeared to be well supported and in good contact with the flanges of the Ijoists
(Refer Plate 22). However, inspection of the floor space on the other side of the joist
showed the joists had been braced by means of a combination of herring-bone and
conventional block noggins (Refer Plate 23). Attempts to test the strength of the herring-bone
strutting showed the bracing could be made to move by hand suggesting it was inadequately
fixed although this could not be confirmed due to difficulties of access.

Examination of the floor void at the site of a second inspection hole cut in an area which had
not been subjected to earlier remedial work showed that intermediary noggins had been
installed between the joists. However some of the noggins appeared not to make close
contact with the joists requiring to be additionally braced against neighbouring noggins by
means of longitudinal braces (Refer Plate 24).

Examination of the same void space in the opposing direction showed the joists had been
braced using a combination of conventional block noggins and herring-bone bracing.
Evidence was found indicating that the block noggins had been cut short meaning they did
not make close contact with the flanges of the joists at both ends (Refer Plate 25).

Reaching through the second inspection hole in Bedroom 3 allowed access to a block noggin
enabling it to be tested by hand for security of fixing. This resulted in the noggin coming away
even before any hand pressure had been applied to it. The noggin had been fixed with a
single annular ring shanked nail inserted at an angle at each end meaning that the depth of
penetration into the flange of the joist had been foreshortened (Refer Plate 26).
The floor decking at the location of the second inspection hole in Bedroom 3 provided further
evidence indicating that the decking was supported by a discontinuous layer of foaming
adhesive rather than the I-joist flange. In some places a layer of adhesive measuring 8mm
provided support for the decking whereas in other places the Caberdek decking was seen to
be unsupported (Refer Plate 27).

Discussion & Conclusions

Suspended Floor

Based on the observational evidence obtained by BM TRADA from the property it is confirmed that the suspended floor to the property did exhibit noise
generation in the form of discernible creaks particularly on the side of the house
corresponding to Bedrooms 3, 5and 1. Noise generation in the form of creaking whilst still
noticeable in Bedrooms 4 & 2, was considered to be less of an issue.
Observations of the floor void in Bedrooms 3, 1, 4 & 2 from the vantage of seven separate
inspection holes revealed two main issues which, together, were principally responsible for
the noise generation which was found. The first of these was that fact that the I-joists were
found to be inadequately and non-uniformly braced by means of intermediary noggins
whereas the second was attributable to the Caberdek decking often not making adequate
contact with the I-joist flanges and being supported mainly by deposits of foaming adhesive
which, themselves, were likely to be discontinuous.

All suspended timber floors exhibit some degree of deflection when subjected to the
magnitude of loading associated with normal footfall and is considered to be entirely
consistent with the behaviour of these types of floor. When such deflection movements occur
in situations where decking panels are used as the walking surface, noise generation in the form of squeaks and creaking can occur unless the decking boards are firmly fixed and fitted both to the joist supports as well as to adjacent boards.
In a scenario where the decking were to be installed in a manner resulting in variable and
non-uniform contact with the joists the opportunity for disproportionate or variable movement
within the decking “skin” is increased with the result that individual panels are likely to
articulate against each other to create the type of creaking noise which was observed at the
site.

The above mechanism of noise generation would be further compounded if not all decking
panels were to be adequately glued along their interlocking joints. This is significant in this
instance given that of the eight inspection holes cut, two instances were found where the
Caberdek joints had not been glued.

If one further considers that the joist layout of the house was such where the flooring to
Bedrooms 3, 5 & 1 (i.e along the side elevation) was supported by slightly longer spans of
joists than those in Bedrooms 4 and 2, the opportunity for creaks to develop as a
consequence of correspondingly greater deflection movement would have been increased.

This gives traction to anecdotal accounts related by the homeowners which suggested
that creaks first began to develop in Bedrooms 3, 5 and 1.

Inspection of the floor voids showed that earlier attempts to reinforce the floor construction
through the insertion of intermediary noggins and herring bone strutting were largely
ineffectual given that noggins were often seen to be cut short or not in close contact with the
joist flanges at either ends. Similarly whenever noggins could be reached by hand it was
found they were fixed by a single nail fixing, and offered little support to the joists as
evidenced by the fact they could be dislodged by hand. This also gives credence to the homeowners view that the remedial measures carried out by the Client, did not significantly lessen the noise generation issues which had occurred initially.

The fact that the floor in the area of the landing and, to a certain extent, the floor in Bedroom
4 immediately adjacent to it, felt springy underfoot is attributable to the absence of any
obvious reinforcement in the form of intermediary noggins which would have helped bind the
floor construction and serve to dampen its responsiveness to foot traffic. Certainly the
Caberdek flooring in both these areas was found to be in relatively good contact with the
joists and strengthens the commentary, already advanced above, relating to the mechanism
of creaking and its relationship with the level of mutual contact between the decking panels
and joists. In this regard the observation that the plasterboard ceiling panel did not make
adequate and uniform contact with the underside of the joist as seen in the inspection hole
cut in Bedroom 4 would explain why the levels of apparent creaking were heard to be greater
in the Dining Room when the floor upstairs was being trafficked than they were within
Bedroom 4 itself.

The fact that the levels of creaking in Bedroom 2 on the opposing side of the landing were
considered to be less severe is attributable largely to the fact that the floor within Bedroom 2
appeared to be reinforced either by way of a reinforcement batten fitted along the lower flange
of an I-joist at approximately a quarter of the span of the room, or otherwise to the presence
of a double joist spanning the central section of the room. Both served to impart greater
rigidity to the floor construction in Bedroom 2.

As a corollary it should be stated that the above observations were based on findings
obtained from eight inspection holes over the area of the suspended floor and therefore the
above commentary is based on a generalised assumption that they are likely to be a fair
representation of the standard of construction occurring over the entire upper floor. Accepting
this, one is drawn to the view that the standard of installation of the floor at the property is inconsistent and, in many instances found, of an inferior quality, which would account for why certain areas of the floor were more prone to creaking than others.

Recommendations

Given the inconsistencies and inadequacies of construction found on the upper floor of the property, the recommendation is made that the Cabedek decking should be removed completely from within all the living spaces of the upper floor and the upper faces
of the I-joist flanges cleaned from all adhesive residues and further checked for any level
disparities between joists.

This will provide the opportunity to brace the entire floor area with new and replacement
noggins ensuring that all block noggins are accurately cut to length (bearing in mind that joist
spacings were found to vary) and firmly fixed with at least two annular ring shanked nails of
adequate length and gauge driven from opposing faces of the noggins to ensure good
resistance to lateral movement from opposing directions. Noggins should be liberally spaced
at 600mm centres and staggered between successive lines of joists to ensure good lateral
bracing to the upper floor structure.

New Caberdek decking should be installed in accordance with the Caberdek installation
guide ensuring that all header joints are supported, and all tongue and grooved joints are
provided with a liberal and continuous application of adhesive to ensure that the entire line
of the joint is bonded.

Site supervision should be provided to ensure that all noggins are secure and in good contact
with joist flanges and have been staggered relative to adjacent lines of joists. Supervision will
also be required to ensure that all installed Caberdek boards are entirely supported by I-joists and noggins and that all contact faces of joist and noggins have been applied with a uniform and continuous bead of adhesive in order to prevent articulation of unglued surfaces when
the floor is trafficked.

There is more to this report regarding the stairs, I will cover this off in a separate blog.

For now here is a short video of one of these noggins:

I think one of the pivotal moments was when the contractor assigned to assist the consultant from TRADA, cut an inspection hole and the noggin fell away to the ceiling below – the Caberdeck board was holding the noggin up rather than the noggin supporting the board.

All of this took place in 2021 – to date nothing has been done to rectify these issues as Bovis have always stated that all the work to the property will be done at the same time. It is now December 2024 and they have gone back on all of these promises, now deciding their only offer will be to buy back our home. 5.5 years of fighting to get our home rectified and once again Bovis backtrack on promises. 5.5. years of constant “inspections”, unnnecessary delays, ineffectual fixes as described in this report, mess, stress, disruption and what feels like a massive weight. I no longer remember what it feels like to live in a normal home.

Have you had similar issues? Please comment below, it would be great to know we are not the only ones going through this.