Last year I made the first version of cold frame in the back of our house which didn’t really work well. Reasons are many: the construction, bad insulation on the sides, cold air draft from the bottom etc. One more time I did something hastily thinking I can “trick” the Finnish climate and make something simpler than it should be. It needed fixing.


The trees/pots are buried into the gravel at the edge of the foundation so the trees should get the thermal energy from 2 sides: the bottom = ground, the back = foundation. Massive concrete should be a very good thermal capacitor as well as ground which should give away some low temperature heat energy.

These things didn’t need fixing.

The side walls are now polyurethane insulation used for saunas (used already in heated winter storage for winter sensitive trees).

The front is under around 45 degree angle, turned to west, made of 4 cm square profiles wooden frame, wrapped into thick transparent nylon. That was done last year already and I wasn’t extremely happy with the result. The nylon tended to bend under the snow and the 4 cm air insulating thickness became quickly less thus reducing insulating characteristic. To fix that I bought the 6 mm polycarbonate twinwall sheets and set it on the outer wall of the wooden frame. I got now 2 layers between outside and the trees. The twinwall sheet and the outer nylon are almost touching each other so practically I can consider this a 3 layer insulation 4.6 cm thick all together. All is transparent so some light can reach the inside which would be important for temperatures above 0 degrees C. Maybe in spring I can measure with a light-meter what  is the percentage of light which penetrates the inside of the cold frame. Now it is too late, all is already closed and set for coming winters agonizing temperatures 🙂

The construction looks like this:

cold frame3



On a branch of a tree in the middle I set one temperature sensor to be able to follow how well the cold frame functions. Under one of the pots I put another sensor but that one is not connected to the cloud (will do next year the same with that one too) so I will follow manually the state of temperature under the trees, where the roots are. Considering I covered the pots with bark chops the temperature there should be dropping slower and higher than outside for most of the time.

The real-time sensors temperature logs can be seen here:

Cold frame winter storage vs. Outdoor

When I first set this new/fixed construction it showed just 1 degree difference between outside and inside the cold frame. I just couldn’t believe how bad it behaved. I looked for cracks, for possible holes where air would come in just to find none of them.

before fixing

Before and after fixing

And then, one day just popped in my mind. The wooden frames upper edge leans on the houses facade (wooden boards) which is of open type (I don’t know how it is in English, Finnish = “tuulettuva“) which means that under the outside wooden boards are 2-3 cm open space through which air can circulate. This type of facade is common in the northern Europe as it keeps dry and keeps the heat inside the house better while outside is wet and cold. So these holes/open channels for air sucked out all the “warm” air inside the cold frame and gave it out at the top of the house. When I realized that I felt very stupid.

After I made a protection so that the holes are not open to the cold frame inside but are still opened for outside air the temperature started behaving very different.

The temperature is now higher than outside but I still need to see how it behaves when it is extremely cold outside before I can decide if this works well for me.

I can’t stop thinking about trying not to cover the roots with the bark chippings. I did it to improve insulation for roots and that is the most important part. But that way I make more resistance for earth heat energy to come upwards and perhaps increase the temperature inside a little bit more. In one way I am afraid to damage the roots, on the other hand perhaps that wouldn’t be a problem if enough of energy can come from below. Something to think about while watching the temperature development 🙂

Good things about this construction
  • I can disassemble it (in summer I don’t want it to be next to the house
  • It has a big transparent side so more light can enter
  • The cover under angle doesn’t gather snow so it can be of lighter construction
Bad things
  • Once I close it it is not so easy to just open it and peak inside or remove a tree
  • More light, a very big question how much light is actually needed. Some yes, there are conifers but anyway I am slightly questioning it.
  • Because the snow doesn’t gather I loose that as an extra insulation when the winter is heavily ongoing

Time will tell how well this is but the solution is definitely better than last year.

Some photos of the present “installation”:


The nylon under the frame is attached to the frame and it serves as air sealing in the bottom.



Improvements for the next year
  • Easily managed openings on both sides so I can make a draft when temperatures are over 0 degree C. Now I do it by removing some of the “sealing” in the bottom and in one corner but would be better I have something more handy
  • Better enclosure for facade ventilation. At the moment it is just a nylon sheet covering it from the bottom and leaving sides open to outside of the cold frame so the ventilation is still functioning but without cold-frame-inside air.
  • Will install another cloud temperature sensor under a pot so I can follow that temperature too.


A must to read on the subject of winter protection for trees:

An excellent page on overwintering bonsai in Finland can be found at Jani’s blog.

Also an article about his cold frame already proven to work very well.