In our first blog piece about Wufi, we discussed why and when Wufi should be used. This piece gives some pointers on how to tell a good report from a bad one.
Wufi is a powerful tool for moisture risk assessment, and it is increasingly used to analyse unusual and novel construction methods and retrofit situations. However, as with any computer simulation, the “rubbish in, rubbish out” rule applies. Additionally, the result graphs require a degree of skill and experience to interpret and to draw conclusions as to the hygrothermal risk of the construction.
So what should you look for in a good hyrgrothermal risk assessment?
The key questions to bear in mind are:
- Do all the inputs make sense?
- Will moisture accumulate anywhere in my construction over time?
- What conditions (temperature and humidity or moisture content) will be created?
- Do problematic combinations of conditions, times and materials arise?
For example, it would be a problem for a piece of timber to be kept above 95% RH or 20% moisture content, and at room temperature for three years; a piece of granite will happily tolerate the same conditions. Between these two extremes there are many shades of grey.
Wufi Analysis: When is a report not a report?
Wufi allows users to produce a ‘report’ on the inputs and outputs automatically. This is a good way of summarising the many assumptions about materials and conditions, and can be used to produce graphs, but on its own does not constitute a full risk assessment. For a simple construction, it would be reasonable to expect a few hundred words discussing and explaining the results. The standard Wufi output is a convenient way to summarise the inputs so it is a good idea to include the standardised report as an appendix.
Numerical quality is an important consideration. This means checking that the calculations have been successful, and reached reasonable levels of accuracy. Comments on the number of “Convergence failures”, “Moisture balance” and “Equilibrium conditions” should be made. The first two will be listed in the standard Wufi outputs if not in the text of a written report. Usually, the simulation should be run for long enough for the moisture levels to reach ‘dynamic equilibrium’, meaning that in the last few years of the simulation, the results should repeat themselves because the weather is the same. Wufi outputs a variety of graphs, and Wufi 6 has a more powerful tool for creating graphs to show specific points. The assessor might extract some key graphs that show enough information to illustrate one or two points, but too much information quickly gets confusing.
Assumptions and inputs
Wufi comes with a database of materials that can be used in simulations. The information Wufi needs is more complex and comprehensive than is typically provided by manufacturers via type-testing and Agreement certificates etc. Users can set up their own materials but care must be taken to ensure the figures are reliable. The database contains many ‘proprietary’ materials, as well as common generic materials such as gypsum plaster board. If the construction uses a material that’s not on the database, the user can either substitute it for a similar one (e.g. mineral wool of similar conductivity and density from a different manufacturer), or modify an existing material. Any such substitutions should be clearly stated, and comments on the impact on the reliability of the assessment made. In some cases, even greater care must be taken. For example, stone and hand-made bricks are naturally quite variable. There are many varieties of brick and stone in the database, but care must be taken to pick an appropriate input. In some cases it is useful to do a Karsten tube test on the masonry in-situ, which is a way of measuring the moisture absorption. This can either guide the appropriate selection of a material, or be used to modify an existing material dataset to create custom input.
BS EN 15026 required simulations to be based on a weather file that approximates a ten-year extreme in terms of temperature. This can be obtained from Meteonorm, for example, but the ten-year extreme element is sometimes overlooked. In any case, the weather file should be for a local location and micro-climate features may be appropriate too (e.g. rural exposed versus sheltered city).
Wufi can add moisture sources in a construction to represent the effects of moisture ingress via wind-driven rain, infiltration and by other types of air movement. All risk assessments should apply some stressing to a construction, ideally in a parametric fashion. Although it may be relevant to present a case with ‘perfect construction’, real buildings are never perfect and will always have at least a small additional moisture load over and above the surface rain wetting and internal moisture diffusion that are inherent in the model. Running a small handful of cases also makes it possible to show how and why a construction accumulates moisture, rather than just if and where, and therefore it can inform design decisions.
“All models are wrong; some models are useful”
In the case of Wufi, some of the physics involved is still not completely understood, and some effects are not represented for example swelling of materials or pooling water in voids. Wufi is good for looking at trends and typical levels. A Wufi graph might say that at 14:00 on the 9th of June, the relative humidity 54mm through the insulation is exactly 63.74%, but it’s clearly absurd to expect exactly that to occur. It might be more reasonable to use a graph to show that the RH is between 60 and 65% for the month of July, and draw a conclusion from that plus a number of other such observations. Spurious precision and claims of inappropriately high accuracy may be indicators of poor understanding of these issues.
Wufi is a one-dimensional tool. It cannot fully represent a complex 2-dimensional situation, for which Wufi 2D would be required. Qualitative comments could be made about likely conditions at 2D junctions, inferred from the 1D results, but in some cases only a 2D simulation will do.
Complimentary information and analysis
It is sometimes useful to pair Wufi with other tools to enhance a report and draw out further conclusions. For example, a 2D psi-value calculation may shed some light on a situation at a 2D junction where Wufi isn’t able to fully model the problem, but Wufi 2D might be a sledge hammer to crack a nut. Other tools that may sit well alongside Wufi could be a ‘simple’ interstitial condensation calculation (as per EN 13788), 2D and 3D thermal bridging calculations, U-value calculations and whole-building analysis using PHPP, SAP or dynamic simulation such as IES.
Wufi is a sophisticated tool for simulating complicated problems, where there is plenty of scope for uncertainty. Conclusions from modelling exercises should always reflect this uncertainty and express degrees of risk rather than absolute predictions. Experienced users have a good understanding of what Wufi can’t do, as well as what it can, and pair this with knowledge of heat and moisture physics as well as building construction to provide an informed judgement about the moisture-safety of a build-up.