[Greenbuilding] Passiv Haus
John Straube
john at buildingscience.com
Sun Jan 30 19:25:18 CST 2011
Bonjour John
I think we are talking past each other. You have made a nice summary, and I dont have a problem with your summary.
My comment earlier this week is that the claim that the PHPP package is somehow world-beating in its performance is not supported by the any reports I have seen, and certainly not by the data of CEPHEUS. It is pretty good for a spreadsheet. I have heard, on this list, and in seminars from Boston to San Francisco, Seattle to Atlanta, that PHPP is a superior and more accurate model than others have available. It is about in line for a monthly bin-model type program.
So, PHPP is nice, good, but not the most accurate or most powerful model.
Numerous programs, like HOT2000 (back in the day HOTCAN) is a monthly bin program that was very seriously validated against cold climate wood frame houses (in Canada) and test cells in the 80's and improved over time for software.
I am using HOT3000 from time to time, but trust HOT2000 more because I know it has been validated.
As I said, I have PHPP and the 100 page guide to its operation.
On 2011-01-30, at 3:32 PM, JD contact (free.fr) wrote:
> Hello John,
>
> I found this in my outbox from last year~
>
>
> PHPP is a monthly steady state model that has been correlated against
> dynamic hourly simulation model Dynbil, of the PaassivHaus Institute and real world
> measurement.
>
> PHPP demo
> http://www.passive-on.org/CD/
>
> A small tutorial of critical elements for the PHPP (Passivhaus
> Planning Package) but note that the package comes with
> a +100 page manual.
> http://www.altompassivhuse.dk/download/kursus/8_Workshop-phpp-day2.PDF
>
> It is significantly better than
> models such as used in the UK, BREDEM, a steady state monthly model
> used for regulatory calculation but without the competance need for
> good dynamic modelling.
> http://www.aecb.net/PDFs/BRE%20response%20to%20Paper%203%20Mar06.pdf
>
> It is not as good as a dynamic model calculated with finesse and
> expertise. And that is a problem as even with good model description
> (one report showed -31% to +13% results) analysis of the
> performance of good energy modelers is probably around +/- 10% * De Montfort University (DMU),
> in the UK did an statistical analyis of cumulative error variability
> of the performance of SER-RES/SUNCODE in the 90s.
>
> Some of the validation done on ESp-r :
> http://strathprints.strath.ac.uk/5819/1/strathprints005819.pdf
>
> Now if you look at real buildings and theoretical energy modelling
> using steady state models you will see for instance in the UK around 30%
> average difference! So in this respect PHPP is much better and
> probably sufficient for the passivhaus type building.
>
> The study of the validity of PHPP with regard to over 100 buildings
> (Cephus project) is here, yes there were a few big outliers but read the
> report most of these were technical setup / human problems :
> http://www.passivehouse.com/07_eng/news/CEPHEUS_final_long.pdf
>
>
> Arguably the new Hot 3000 canadian modeller using the ESP-r calulation
> motor developed in Europe by Strathclyde University (and the European
> reference and now open source) should be quite good. It is a 3d
> volumetric model with heat balance equations being solved at every
> time step by finite difference method. In its basic form it is not for the
> faint hearted, it being a heavyweight university research tool.
>
> ESP-r - Heat balance equations are solved with the finite difference
> method at each time step.
>
> energyplus - Heat balance equations are solved with the response factor method
> http://wiki.naturalfrequency.com/wiki/Thermal_Analysis_Methods
>
>
> In most dynamic models in the US the heat balance equations are solved
> with the response factor method models using algorithms (approximations) fourrier
> transformations to shorten calculation time. It is a
> little easer to write algorithms for service plant... thus the bias
> in America which extensively uses HVAC. It is very good but arguably
> the ESP-r model is a
> little bit more accurate, particularly calulating the free running
> performance of heavy weight buildings.
>
> You can see the comparatif performance (not against real buildings) of the major modellers in the
> Besttest standardised performance test.
> http://www.nrel.gov/buildings/energy_analysis.html#bestest
>
> There was a program of validation of software against controlled test
> cells (one room) in the European Union during the 80-90s and as I
> remember ESP-r was the best but certainly not perfect.
>
> In the real world you have people and HVAC controls which do
> the silliest of things sometimes!!
>
> regards
>
> John Daglish
>
> Friday, July 16, 2010, 5:58:44 PM, you wrote:
>
> JS> I think that is the study that is often mentioned. Median tracked but variation was large, eg, there is a table in the report (which mostly deals with multifamily, less so with single family) and
> JS> the ratio of the modeled energy prediction for HEATING only divided by actual varied from about 0.5 to 1.7. Quite the scatter!!! Also, this is just for heating energy, which should be pretty
> JS> easy to get. Most standard models would actually do better.That said, low energy houses like PH have a lot more noise, as it takes little in terms of solar reflections, higher interior electrical
> JS> gains, small mechanical system or enclosure flaws to make a big difference.
>
> JS> But, the conclusion I reached is the PHPP is no better, and probably a bit worse, than most models. Yet I keep hearing about this study that shows PHPP is great. Have asked numerous people where
> JS> this study is .....
>
> JS> I still LIKE PHPP and would use it or similar to do modeling for low energy houses. But I just cannot find any reason why it is better.
>
> JS> On 2010-07-16, at 11:42 AM, Alan Abrams wrote:
>
>>> John
>>>
>>> sorry, it's not at my fingertips; it was presented during consultant class. the 115 house sample sounds familiar; it may be the same one you refer to. IIRC, the correlation was in the 80-85%
>>> range (of very low numbers to begin with), but the median tracked right on the money.
>>>
>>> still, memory plays tricks; will dig for it when I get a chance...
>>>
>>> AA
>>>
>>> On Fri, Jul 16, 2010 at 11:09 AM, John Straube <john at buildingscience.com> wrote:
>>> Can you tell me where I can find this study?
>>> The only study I have found is the CEPEUS 115 home study and it did NOT show very good correlation. Adequate yes but not good.
>>>
>
> JS> Dr John Straube, P.Eng.
> JS> Building Science Corporation
> JS> Somerville MA Waterloo ON
> JS> www.buildingscience.com
>
>
>
> JS> _______________________________________________
> JS> Greenbuilding email list
> JS> Environmentally-preferable design, construction, building elements
> JS> List info: http://listserv.repp.org/mailman/listinfo/greenbuilding_listserv.repp.org
> JS> List email: Greenbuilding at listserv.repp.org
> JS> Managed by BuildingGreen, Inc. http://www.buildinggreen.com
> JS> publisher of Environmental Building News and GreenSpec
> JS> Hosted and archived by REPP / CREST http://www.crest.org
> JS> To get on the list:
> JS> http://listserv.repp.org/mailman/listinfo/greenbuilding_listserv.repp.org
> JS> or mailto:greenbuilding-request at listserv.repp.org?subject=unsubscribe
> JS> To get off the list:
> JS> http://listserv.repp.org/mailman/listinfo/greenbuilding_listserv.repp.org
> JS> or mailto:greenbuilding-request at listserv.repp.org?subject=unsubscribe
>
>
>
> --
> Best regards,
> JD mailto:johndaglish at free.fr
Dr John Straube, P.Eng.
Building Science Corporation
Somerville MA Waterloo ON
www.buildingscience.com
More information about the Greenbuilding
mailing list