Is SHGC a very rough estimation of solar gains?

Hi there,

I am trying to understand how SHGC is impacting solar heat gains. I am working on a simplistic model that is a rectangular box with one west-facing window. There are also 2 shades to account for contextual masks.

EDIT

I’ve probably got some answers from this post : Where can I find a Glazing Conduction Definition? - #5 by erydjunaedy. However I feel my post stil has some interest. So I’ve pushed it on the forum !

Study

  1. I made a radiation study with the LB-Incident-Radiation component and found a 2719kWh incident sun energy on the window (external) surface.

  2. I made a radiation study with the HB-Incident-Radiation component and found a 2789kWh incident sun energy on the windows (external) surface. The relative error is only 2.6% between LB/HB which I suppose is quite ok. This study allows me also to investigate further monthly results and contribution from direct / diffuse sun.

  3. Finally, I made a HB-E+ study. I am using a “simple window material” with U, SHGC and T factors. I have run this simulation for different values of these factors.

Remarks

  • U and T values don’t have a significant impact on the ratio of (E+ Solar Heat Gains)/(Total Incoming Sun Energy).
  • This ratio is clearly dependent of the SHGC (when SHGC increases, the ratio increases, when SHGC decreases, the ratio decreases as expected).
  • However the ratio is not proportional to SHGC (for instance varying SHGC from 0.5 to 0.25 leads to a 66% loss in heat gain).
  • With U=12, SHGC=0.99, T=1 solar gains are 2368kWh (87% of the incoming sun energy)
  • With U=0, SHGC=0.99, T=1 solar gains are reduced to 2152kWh (79% of the incoming sun energy)

This field describes the value for SHGC, or solar heat gain coefficient. There are no units. This is the rated (NFRC) value for SHGC under summer cooling conditions and represents SHGC for normal incidence and vertical orientation.

Questions

  1. Would you confirm that SHGC just roughly accounts for solar gains through a window ? Thus allowing to compare various glass materials but without a quantitative information about how much heat goes in for a given sun radiation (HG = SHGC x Radiation ) ?

  2. I understand that, internally, E+ conducts it’s own radiation computation. Is it possible to output the results of this computation and access for instance the hourly amount of sun power a face/window receives (and diffuse vs. direct) ? What would be the correct output to request ?

Hi @lionpeloux, I can answer to #1 is SHGC accounts for solar gains - that is exactly it.

The u-value of the glass only considers the the heat gain from non-solar gains (such as heat from external air and its current temperature). Hence the resistance i.e m2K/W of the material (whether glass pane or cavity) impacts the u-value.

The SHGC or g-value is impacted by the thermal transmittance value of the glass only, especially the outer pane (if it is double or triple glazed). so a SHGC of 0.5 means that 50% of the solar heat is transmitted from the sun rays.

You can always run a peak load analysis to understand the impact of the glazing only by cycling through the results. I find it to be a much faster exercise instead of running energy simulations when its a matter of understanding the impact of a specific design intervention

You mentioned the ratio of reduction not being proportional, thats because you have to also consider delay in peak temperatures within the outside and inside. The heat from outside will always make its way inside, but providing a good glazing system will allow for a “lag” to develop. For example if at peak the sun energy is 2500 kWh, but because the SHGC is low, the 2,500 will be “delayed” in transferring heat into the space, hence reducing energy/heat gains by a non-proportional factor.

As for the second question - I am sure someone can provide a better answer, I usually use the below to cycle through the different faces to check their energy flow:

You can use the HB Read Result Dictionary to see the full list of all EnergyPlus outputs that you can request for a given model. Then, you can request those outputs from the simulation using the HB Custom Simulation Output component, and parse the results into Grasshopper using the HB Read Custom Result component.

I think the output that shows you the directly-transmitted solar energy is Surface Window Transmitted Solar Radiation Energy but this will be different if you want the incident solar radiation or the portion of the solar heat that conducts through the window.

Thank you @hmurya and @chris. This is really helpful for me to dive deeper into this topic. I wasn’t aware of the HB Read Result Dictionary … worth knowing it :slight_smile:

However I didn’t find any output that would correspond to the “raw solar energy falling on the window outside face” to compare with my radiance study.

Average hourly profil by month shows well when the window loss/gain heat during a day (as suggested by @hmurya ).

Hopefully this is a useful addition to the conversation. My understanding of SHGC and g-value is that they are measures of solar gain when radiation is normal to the glazing (or some kind of average? I’m not certain on the details). In reality angle of incidence will vary the solar gain through a window depending on the transmittance properties and radiative properties. In IES I’ve seen two glazing constructions which report the same g-value give significantly different solar gain results, particularly when comparing double and triple glazing.

Anecdotally, I think if you use something like WINDOW you might be able to get a better breakdown of how solar transmittance varies by angle for a given construction, but I’m less sure how you could then also look at the variation in radiative effects. I’m sure my terminology is all over the place but hopefully it opens the path for more research.

That said, to answer the title, I believe it is a rough estimation of solar gain, but only rough and there’s more to it.

Here’s a snippet from Desiging Buildings on the topic.

“Actual solar heat gain is dependent on the angle of incidence of solar radiation on the glazing (and so the proportion of diffuse and direct beam solar radiation) as wellas the spectral make up of the solar radiation. However, as a simplified method, manufacturers will often only provide a spectrally-averaged solar heat gain coefficient for normally-incident solar radiation. Values for other angles of incidence and for diffuse solar radiation can then be estimated using standard equations or tables for similar windows.”

Thanks @charlie.brooker.
For sure this is a useful addition !

I realise that my initial intent was “just” to evaluate, on a real-life case, the sun power that a certain window received and what kind of action I should take to prevent over heating in summer (france).

What I’ve learned so far :

  • it’s a tough topic !
  • SHGC really gives you only a rough idea of a glass/window behaviour regarding overall solar heat gain. So it can help you choose between various glasses on a “catalog”.
  • SHGC will not help you to quantify solar heat gain/loss through your window. The equation HG = SHGC * Global Incident Radiation is just too simplistic.
  • Diffuse radiation can be responsible for quite a large amount of heat gain (I wasn’t expecting that before starting this study). In my case (france) this is about 50% (see chart below). So blocking only direct sun with a shade will roughly act on half of the problem (of course day light is also of interest …).

3 Likes

@lionpeloux that’s great stuff !

Just for my reference, did you specify those variables as manual HB Custom Simulation Output, as these do not come out by default from the HB_readEPSrfResult component.

Hi @andrea.botti , it’s been a while I did this study, and I indeed specified this through the Custom Simulation Output

The solar factor g (same as SHGC) is the ratio of the incoming solar energy (sum of the energy passed directly to the interior plus that absorbed by the glass panes and transmitted back to the interior by radiation in the long-infrared spectrum) to the incident solar energy.
This calculated in the direction normal to the glass according to standard (I seem to remember Uni Iso 410?).
However, the behavior of glass, as correctly said by some in previous posts, is dependent on the angle of solar incidence.
In Energyplus it is possible to use the BSDF calculation method via complex fenestration states, a 145x145 matrix simulated for each layer of our insulating glass unit and also for shading! (This way we can simulate non-standard shading, like as expanded metal sheets or hand-made design, by correctly considering reflections to the glazing itself and transmission as a function of solar incidence angles.)
Then if you want, you can derive the value of g for each instant of calculation and/or evaluate it statistically for periods of time.
The variables in E+ are:
.Surface Window Transmitted Solar Radiation Rate
.Surface Outside Face Incident Solar Radiation Rate per Area.
.Surface Window Inside Face Glazing Net Infrared Heat Transfer Rate (note that for the latter you must add the following string in E+ to enable the advanced variables: Output:Diagnostics,
DisplayAdvancedReportVariables;)
It is difficult to explain all the processes, I would like to make a video-tutorial but I never have the time…
I’ll leave you with a few example slides, based on the input of a selective insulating glass unit that by data sheet had a g-factor of 0.26:

here an example of the result with the BSDF also calculated for the expanded metal sheet )beyond the double glazing:

The BSDF matrices of the glass panes, layer by layer, are directly exportable from LBNL Window and imported into directly into E+ for calculation.
The BDSF matrices for shading is created by GenBSDF (radiance) and then imported inverted into Window…
In terms of energy, the zone inputs and consequently the calculated energy change a lot.
Good work!
See you soon
Massimiliano

(Sorry for my bad english…)

2 Likes