Honeybee energy consumption simulation error with Galapagos


I made a model with variable insulation thickness. I have six variables: east wall insulation thickness, west wall insulation thickness, south wall insulation thickness, north wall insulation thickness, roof insulation thickness and floor insulation thickness.

When the thickness change, annual heating load /㎡and annual cooling load /㎡change.

I would like to use Galapagos optimize the six variables for lowest full life cycle cost.

Full life cycle cost is related to two aspects:cost of insulation material and cost of building operation. I built a function to represent it.

Here are my questions:

Galapagos works fine, but annual heating load and annual cooling load look wrong, they are obvious smaller than the ones without Galapagos.

When I am working with all of the data, I find that when the thickness change, heating load/㎡ and annual cooling load/㎡remain stable, and stable at a lower value.

If I remove Galapagos, load and annual cooling load/m2look normal.

The same parameter produces different results. For example:when east wall insulation thickness=0.06, west wall insulation thickness=0.06, south wall insulation thickness=0.06, north wall insulation thickness=0.06,roof insulation thickness=0.07, floor insulation thickness=0.11,there are two results:

①cooling load/㎡: 57kWh/(㎡·a), heating load/㎡: 87kWh/(㎡·a)

②cooling load/㎡: 54kWh/(㎡·a), heating load/㎡: 64kWh/(㎡·a)

I can not attach the gh file because I am new users. I’ll appreciate it if you give me a hand.

Thanks in advance!

Why don’t you add drive folder link?


Thanks for your advice, sorry for the late reply. This is my link:


Thanks in advance!


I was looking your woklow, it’s hard to understand your logic but, especially your formula.

Can you explain more this, how same parameters can give different results? When I run on my computer, taking same results.

By the way, are you sure connect right inputs to your formula, I mean cooling → b, heating → a ?


Thank you very much for your reply. My formula is related to full life cycle cost. It is a little complex, but I think it should be right.

I’ve been running for 35 generations last time, I find that the optimal solution of every generation looks strange, when the thickness change, heating load/m2 and annual cooling load/m2remain stable. like this,
But I find maybe there’s something wrong here. Each name of EPMaterial is pu. That means I have six pu in this file. I think it might have an impact. But I am not sure.

I am trying now. I will reply as soon as I have a result.

Thanks again for you kind help.


It’s really weird, my simulations looks really reasonable as you can see results below;

Can you try again with my .gh ? I changed some small things.

0423simianqiangtibutong.gh (629.4 KB)

I will try again with your .gh. It looks reasonable. Thank you very much! :grinning: :grinning:

I changed every EPMaterial ’s name and was really delighted to find the results look reasonable now. To test this further, I checked with your file, they are the same. Here is some of the results.

I really appreciate your help! And sorry for my careless. I hope there are more opportunities after the exchange of learning with you! :grin: :grin: :grin:

It’s normal, no worries.

You have already solved your own problem, nice job.
Take care, be safe!