Re: [EnergyPlus_Support] Doubt about Airflow Network and leakage through the envelope

Hopefully I can help you a little, but you may need to ask your question on a DesignBuilder user group.

What has been setup is a typical airflow network, and you may have limited control over what is going on if you do not completely understand the concepts of AirflowNetwork computations. The three AirflowNetwork objects define the surfaces involved in the calculation, the type of crack defined for each surface, and the baseline air conditions for the flow calculation.

Apparently DesignBuilder characterizes leakage for various construction air tightness for you. I can tell you that a typical n is about 0.65 and it will vary with the nature of the actual airflow path your building has (I do a fair amount of whole and partial building air testing). n canâ??t be less than .5 (a sharp edge orifice) or greater than 1.0 (laminar flow through a tube). C is doing your volume work, and DesignBuilder is giving you a baseline C for 1 Pa pressure based on the tightness of the construction. That baseline C is modified by the shape of the surface. The baseline C is referenced to a unit length of crack. Each surface will be assigned some crack condition, for instance it could be a perimeter crack around a window. The Crack Factor is a multiplier to C that depends on the geometry and calculation procedure applied to the geometry. Neither C nor n varies with the pressure. Since E+ needs mass flow, the air density has to be adjusted based on air conditions at the time of flow (the viscosity of air does not change that much over the typical building temperature range so the volume flow is constant even though the density changes).

Now all that is needed is the pressure on each crack. E+ will calculate stack pressure (warm air rising, and you need some inter-zone openings defined to link individual zones). If you have a simple rectangular building, E+ will calculate the wind pressure coefficients for you based on wind direction and location of the surface. If you have a complex shape, you should use CFD to generate wind pressure coefficients for multiple wind directions. Now you have all the information to solve a series of linked pressure/flow equations for a pressure balance.

Now comes the questionable part of the solution, your initial assumptions and how you link zones. DesignBuilder evenly distributes the baseline leakage to all surfaces, but that may not be the actual condition of your building. Suppose that windows leak more than walls, you should have a separate description of window leakage to reflect the difference. A wood framed wall leaks more air than a roof with a full single ply roof membrane, so the roof air leakage should be characterized differently. (In one whole building air test we performed, we were able to show that three different zones of the building has different air leakage rates.) If you have a middle of a facade zone defined with no connection to adjacent zones, you could wind up with air flow calculated by stack pressure alone because wind pressure will not be transferred to adjacent spaces, so there is no differential pressure to drive airflow.

Read the engineering reference for how you crack length are calculated and how windows are treated. Think about your model and what you expect the flow paths will be.

We have been modeling buildings that we have air tested and have energy use data for a couple of years. We will be calibrating our models and reviewing the accuracy of modeling and the model assumptions. We will likely publish and present our findings within the next year or so (it is an overhead project and paid work take preference for our time).

Ned Lyon, P.E. (MA, WV)
Staff Consultant

SIMPSON GUMPERTZ & HEGER
781.907.9000 main
781.907.9350 direct
617.285.2162 mobile
781.907.9009 fax
www.sgh.com

From: EnergyPlus_Support@xxxxxxxxxxxxxxx [mailto:EnergyPlus_Support@xxxxxxxxxxxxxxx] On Behalf Of m.aldegheri@xxxxxxxxx
Sent: Friday, February 14, 2014 6:21 AM
To: EnergyPlus_Support@xxxxxxxxxxxxxxx
Subject: [EnergyPlus_Support] Doubt about Airflow Network and leakage through the envelope

Dear all,

I have a question about the calculation of the air mass flow leakage through the envelope.

My building have:
-a natural ventilation system (windows are opened with a control system)
-a mechanical ventilation system which provides a define airflow rate (ach/h)
-infiltration through the envelope discontinuity (I don't have particular vents or holes on the building, but I imagine that I have anyway a leakage because of the envelope is not completely airtight and there could be some cracks in wall joints etc..., please correct me if I'm wrong...!!!)

I have many doubt about the infiltration calculations.
I'm creating my model with Design Builder and (because of some aspects I will not report here now) I have to use the "natural ventilation calculated" in the model option.

Under this option, referring to the airtightness of the building, DB let me choose a level of airtightness (from poor to excellent). Each level is associated with a defined couple of value ("C" and "n") which are used to model the airflow mass according to the equation Q=C(dP)^n (reported in the manual) and there is a characteristic couple for each type of internal or external surface of the building.
The online tutorial explains that for new buildings the level "good" can be chosen.
Exporting the .idf end reading it with E+ I notice that this procedure creates these objects:
-AirflowNetwork:Multizone:Surface
-AirflowNetwork:Multizone:ReferenceCrackConditions
-AirflowNetwork:Multizone:Surface:Crack

Reading the I/O reference (or engineering reference I don't remember) manuals the air flow mass is calculated by
Q=(Crack Factor)*Ct*Cq*(dP)^n
where:
-Crack factor is not specified in the manual
-Ct is a corrective factor by temperature
-Cq is the air mass flow coefficient in kg/s-Pa^n @1 Pa

My question is: how can I be sure about the correct adoption of C and n value in Db and its use in E+? The choice of the "good" level of airtightness seems to be blind action.
I try to explain better, I know that the the best way to evaluate infiltrations is through a blower door test, but the building I'm analyzing is under construction so this way is excluded. I have been suggested to check the parameters used in E+ with the ones reported in EN13465:2004. I red this document but I couldn't fine the clear definition of the parameters I'm looking for.

First of all, is my reasoning correct?
Am I calculating the loss through infiltration correctly?
Is this way is correct, how can I have a confirm of the correct adoption of the coefficient used in E+?

Can anyone help me? Maybe there is anything that I didn't understand?

Thank you all,
Marco