What's New

View | 2008 | 2007 | 2006 | 2005 | 2004

JULY 05 - PARTNERS IN INNOVATION REPORT ON IMPROVED THERMAL AND MOISTURE PERFORMANCE OF PITCHED ROOFS NOW AVAILABLE

Introduction

The Partners in Innovation project that was set up in September 2001 to investigate matters concerning improved thermal and moisture performance of pitched roofs has now published its report. The work principally concerned itself with investigating 'cold' pitched roofs, i.e. those with insulation on a horizontal ceiling with an accessible loft above. There were two reasons for this:

• 1. the issues surrounding cold pitched roofs were complex and controversial and therefore demanded an considerable degree of study;

• 2. the large amount of work carried out in other countries, especially at the University of Leuven in Belgium, on 'warm' pitched roofs, meant that they were well understood and any further work would be in danger of reinventing the wheel.

A range of work was done during the project which is summarised below;

East Kilbride Test Rig

Two sets of four roofs were monitored in detail over the winters of 2001 - 2002 and 2002 - 2003. In the first winter, a comparison of a conventional ventilated 1F felt roof with three variants of 'unventilated' roofs with low resitance underlays (Tyvek 2001B Pro, with a quoted vapour resistance 0.16 MNs/g) and different combinations of counterbattens and batten space ventilation, demonstrated small differences between the systems, The rafter moisture contents in one of the roofs, with lapped but not taped Tyvek, were consistently higher that the others, but within safe limits; this roof was also more prone to condensation. The air tightness of the four roofs was not measured at this stage.

A more detailed study in the second winter compared the ventilated roof with 1F underlay with roofs with Tyvek 2001B Pro as before and two combinations of Roofshield, which is air permeable and has a quoted vapour resistance of 0.09 MNs/g. In both winters, it was found to be extremely difficult to eliminate airflows into the 'unventilated' roofs, a reduction of loft ventilation rate to between 12% and 30% of the fully ventilated loft, was the best achievable. Regression analysis of the results from 2002 -2003 allowed the four roofs to be compared under standard environmental conditions and demonstrated that the risk of problems was not very different between the four roofs, with some small differences in performance:

• The ventilated roof with impermeable underlay was marginally colder with slightly higher relative humidities than the unventilated roofs with permeable membranes.

• The theoretical risk of condensation, as predicted from the difference between the measured loft dewpoint and underlay temperature, was lower in the 'unventilated' roofs with LR underlays, with a resistance 0.16 MN.s/g or less, compared to the ventilated loft with the HR underlay and the LR underlay with a resistance of 0.09MNs/g, which was also air permeable.

• Although the theoretical risk of condensation was higher in the roofs with air permeable LR underlays and with the impermeable HR underlay, compared to that in the roof with air impermeable LR underlay, the actual incidence of condensation was less. This is consistent with air permeability or ventilation giving additional protection against condensation

Reports of condensation

Over the winter of 2002 - 2003, there were a number of complaints of condensation in the lofts of newly completed houses, with an LR underlay and no designed ventilation. Although complaints of condensation is an imprecise, subjective measure, the number of complaints received on each day, was strongly associated with the daily minimum temperature, with a peak of complaints during a cold spell when minimum temperatures fell below -4°C for a number of days.

Drying of houses

It is assumed that the extra moisture load caused by 'wet' materials drying out in new houses, leads to an increased risk of condensation in the loft. Wet trades will impose some extra moisture load on houses; however there is little evidence of the size of this effect. The moisture produced within a house by the normal activities of the occupants can vary from 5 kg/day for a pensioner living alone, up to 20 kg/day for a family with a number of small children. If the drying of the fabric adds something of the order of 1 - 2 kg/day to this, it may not be important, however, if it adds 5 - 10 kg/day it is obviously highly important. Data is needed on the actual rates of evaporation in new houses.

Air leakage though roof coverings

At the start of the PiI project it was assumed that air ingress through many tiling systems would be so limited that it would be necessary to install dedicated batten space ventilators, if an LR membrane was used with an unventilated loft. The default situation, reflected in BS5250:2002, was 'you must install batten space ventilators unless you have evidence that the tiles/slates are very air open'. Quantitative pressurisation tests and qualitative smoke tests carried out in a number of roofs have both demonstrated substantial air leakage though all the tiling systems tested. It is now felt that the situation can be summed up as : 'It can be assumed that most natural slating and tiling systems are sufficiently air open, unless there is evidence that they are air tight. Fibre cement slates, metal tiles or sheet coverings would be considered insufficiently air open''. It has been possible to define an objective test for air openness of tiling systems based on BS5534, Annex L, but with equipment to measure flows down to 2Pa.

Air leakage around loft hatches

Tests of air leakage around five commercially available loft hatches, all marketed as air tight, showed a variation of 50 to 1 between the best and worst hatches, with no guidance available to aid designers/builders which is the appropriate one to choose.

Energy Saving

One of the justifications for the use of 'unventilated' lofts with LR underlays is the potential energy savings from restricting ventilation in the loft. Claims have been made up to 25% energy saving. The evidence for this comes from some research done in Finland and one pair of houses at BRE Garston; the data from the latter was inconclusive in cold weather. One study of cold roof performance in Finland showed an energy saving of 1 - 2 % when loft ventilation was eliminated. When air was sucked down from the loft into the house, using the ceiling insulation as a heat exchanger, much greater savings results, however this is hardly a realistic situation.

A simplified steady state model, which uses the air flow equations developed for RoofCond (see below), has been developed to assess the parameters that affect heat flows and energy use houses as the parameters of the roof are changed. This suggests that it is more likely that any savings will come from improving the airtightness of the ceiling than from reducing air flows through the loft.

RoofCond Software

The software package RoofCond that was developed during the course of the project was designed to carry out sensitivity analysis of the factors that effect condensation in cold pitched roofs. Further work is needed to validate its use; in particular it should be compared with measured data in buildings. However when used for sensitivity analysis the following points can be regarded as reasonably robust:

• Lower ventilation rates are needed in a LR roof to give the same level of protection from condensation as in a fully ventilated HR roof, all other parameters being equal.

• As a comparative sensitivity analysis tool it can be used to compare the condensation risk of a conventionally ventilated roof with a HR underlay directly with a roof with an LR underlay.

• As the ventilation in the loft is reduced, the condensate peak rises with both underlay types.

• As leakage through the ceiling is reduced, the condensate peak falls with both underlay types.

• As the vapour resistance of an underlay is reduced below 0.25 MNs/g, the condensate peak falls.

• Addition of ridge ventilators lowers the condensation peak.

All these trends can be considered to be representative of the performance of roofs in practice.

Click here to download the full Pii report PDF >

Back >