Sigaldry

As with anything in life, your mileage will vary sadly, depending on who does your calculations, how conscientious they are, how good the tool is they use and the quality of the information that they receive. Costs can vary significantly, some overcharge (in my opinion), some undercharge and quality may suffer as a result (don't get me started on RDSAP/DEA produced EPC's where the assessor has to travel an hour each way, do a survey, write it up with all necessary supporting information and gets offered a pittance for doing so - the EPC will be worth very little in that circumstance and even with auditing from accreditation schemes that's an issue. I posted a good while ago on greenbuilding forum that the latest version of SAP is very much fit for purpose, but that its purpose is assessing compliance with the energy and carbon emission requirements of the Building regulations and for providing a methodology for comparison of performance of dwellings (assuming the compared dwellings are occupied in accordance with the assumptions that SAP is based on). Those assumptions have a purpose, because we (people) aren't the same - so one person will heat a home 24 hours a day at 26 degrees - some will even have a good reason for that, others will heat to the minimum. The same house could be occupied by one person, a couple, a small or larger family and will do so in differing ways. SAP and the EPC allows you to compare two similar homes and gain a general understanding of their comparative energy efficiency, running costs and carbon emissions from heating, hot water, lighting etc - it's not perfect, but it's not awful. The Zero Carbon Hub’s review of modelling tools in 2010 noted that no modelling tool can perfectly represent the real world, but overall concluded that SAP actually compared well with other modelling tools available (and that included PHPP). The key issue I feel, is whether the inputs that go into an as-built assessment can be relied upon - which is a whole other matter. Better validation of inputs is needed. Through the build process, there needs to be more evidence and feedback built in whenever changes might occur to capture these / ensure compensation for poorer performance inputs is made. Accommodation of quality assured processes needs to be considered. The data that goes into an EPC is assessed / audited by an accreditation scheme and assessors will get pulled up if there are deviances. Is it perfect, no. Saying that anyone can do a SAP is a misnomer - anyone competent can do so, many people aren't however competent to do so. Equally anyone could in theory do a U-value calculation, the number that can do so competently is somewhat smaller (following all appropriate conventions, standards and using the right values for materials). A good energy assessor will not make assumptions (or only very few, they will deal in certainties and evidence and explain the issues and complexities, they will understand thermal bridging and junction detailing, have an understanding of heating and hot water systems, ventilation requirements and systems, photovoltaics, solar thermal etc, etc - but not all will be competent and even the most competent in life can make mistakes. Competency costs - but so does incompetency. You also need to be clear that an energy assessment will be very unlikely to ever match how a dwelling actually performs in use, once real occupancy and building use comes into play. The SAP is done at design stage before you start work (or it should be); If you propose changing things, it should be updated and checked before doing them and at the as-built stage for a completion certificate, which BCB wont give without the EPC (usually), there shouldn't be any assumptions, just known's, which the customer and builder should know and be able to evidence - a site visit would be lovely, but almost all on construction assessors don't do them (some will), but you'll pay for it and those assessors are mostly general surveyors/architects who also offer other services. If there's sufficient evidence, a site visit won't necessarily be any better or worse than well evidenced documentation. You often can't see how something went together once it's been done anyway. Garbage in = Garbage out (GIGO); If you enter rubbish and construct rubbish, you'll get rubbish.