In winter, bees consume their stored resources for food and heat. Depending on the latitudes, the winter can be more or less mild and the colonies can go through a variety of situations. In which case you should consider hive insulation?

The month of January 2021 was particularly fresh in the south-west of France. Days with 2 or 3°C maximum temperature had been usual. I was fearing my bees being stressed under those conditions. They already had to overcome the hornet season in autumn and now they were facing unexpectedly cold days.

To put this situation into perspective, during these last few years, I did not had any worries with the winter crossing, but this month of January, there are two things that make me wonder that some decisions should be taken.

The first one is the low temperature alerts I received several days in a row. 4°C or less reached inside the hive, that's cold!

low temperature alerts inside the hive
low temperature alerts inside the hive

The second one, which definitely motivated me to act, is the observations after a snowfall. I realized an interesting difference on the roofs of the hives between the R6 which is empty and the others inhabited with bees that I could clearly hear "vibrating".

The snow completely covers the roof of the R6 hive, while it has melted all around the center stone of the other hives. Isn't this evidence that the bees are heating the inside of the hive and that some heat is being dissipated from above?

Hive without insulation
Hives without insulation.

Based on this observation, I wondered what the heat loss through the roof was. Clearly, if there is one side where the losses are important, it is the roof.

It was a long time I hadn’t taken out my old heat transfer equations. I used them for many years in sizing helicopter turbines. Here comes the opportunity to dust them off and to use them for beekeeping.

Problem statement and assumptions

To state the problem, we start from a few principles:

We know that bees aggregate in a cluster. The minimum external temperature of this cluster is 7°C. Below that threshold bees collapse:

The cluster aggregates as soon as the outside temperature drops below 15°C; it is complete - i.e. all the bees in the colony have joined it - at 7°C; the individual rapidly collapses at lower temperatures. It is generally in the shape of a sphere or ellipsoid, sometimes flattened on one side when it is, for example, just under the frame cover. by Janine KIEVITS, Abeilles & Cie n°131 4-2019

We can therefore conclude that the cluster can be modeled by a sphere and that the temperature of this boundary is 7°C in the worst case. This sphere being in a cooler environment, it loses heat by convection towards the roof. Warm air always rises, even within a hive!

Then there is the roof. In my case, feeder and roof, are all made of wood. The thicknesses are around 1.5 cm. The tray is made of 5mm plywood. The warm air that rises from the cluster heats the ceiling and is transmitted by conduction/convection through the roof.

There are some classic equations to calculate that kind of heat transfer configuration. It takes a bit of time, but it’s quite straightforward. You can find the formulas here, or here for example.

schematic of the problem thermal hive

The equivalent thermal network is displayed in the following diagram. In this model it is assumed that 100% of heat leaks from the top. This seems a bit of a strong approach - but (at our latitudes) wax frames avoid transverse losses quite well - and as mentioned before, the heat always rises. We will verify further on if this hypothesis holds (it even holds very well!).

The resolution of the system of equations provides the heat flux dissipated (in Watt) that varies between 0.04 and 2W depending on the outside temperature (+7 to -20°C). We are now fixed about the order of magnitude: The colony permanently dissipates 1/5th of a low-energy light bulb.

Now we would like to insulate the hives. What would be the effect of such an action?

Insulate the hives with 4cm polystyrene

Consider adding a 4cm extruded polystyrene block under the hive roof (count an investment of 3.95€ per plate for 3 hives at your preferred store). We add this constraint in our thermal model and recalculate the dissipated powers.

With the insulated hive, the dissipated powers are much lower: from 0.02 to 0.86 W for the same outside temperatures. The insulation of the roof - at these temperatures - therefore has a significant impact!

Table sumarising power dissipation
Dissipated power as a function of ambient temperature and expected gain with 40mm roof insulation

Quantity of honey consumed by bees

As said above, more heat means more honey consumed. We can therefore push this exercise a little further, by calculating the amount of honey consumed with and without hive insulation. Be careful! Here we consider only the honey consumed for heating purposes, not for feeding. In other words, the actual honey consumption by the bees must be higher than the values we will find.

Honey contains an energy of 13 kJ/gr. It is this energy that the bees dissipate in the middle of the cluster to warm all their fellow bees. It is by the movement of their muscles that they produce this heat, and since any exercise is a source of fatigue, a rotation is established between the bees of the cluster boundary and those at the core. They shift positions so that the bees in the center rest and those in the periphery come to work and warm up a little too!

Converting power needs into honey needs, here is what we find :

graph comparing honey consumption with and without insulation
Honey consumption for heating as a function of ambient temperature, insulated (green) and without insulation (red).

The effect of the insulation of the hive has a very noticeable impact on the consumption of honey (for heating) which decreases considerably. At an external temperature of 0°C the consumption decreases from 3gr/day to 1.5gr/day. The saving is 50%!

In view of these results, I folded the computer to go straight to the store! And here is the work, cut out to the pattern and installed in a blink of an eye 😃

Hive insulation
Cutting of 40mm polystyrene boards and installation between the roof and the feeder.

Some routine checks

To bring this scientific approach until its end, once the model is built, it is a good idea to validate it against actual measurements. As each hive is equipped with a temperature sensor and the outside temperature is also measured by the scales, I have retrieved some measurements over a period of about 10 days.

Sequence of internal/external temperatures measured over the period.
Sequence of internal/external temperatures measured over the period.

The graph below shows the correlation between internal and external temperatures. The blue dots represent the measurements and the lines correspond to the models described above with and without insulation.

representation of the two thermal models with and without insulation and correlation with the measurements
Graphical representation of the two thermal models with and without insulation and correlation with the measurements (blue point cloud) before insulation. The model is representative of reality.

It can be seen that the model without insulation (red) correlates quite well with the measurements. I admit that it fits much better than expected!

There is also an intrinsic limit to the model for ambient temperatures above 7°C. This can be explained by the assumption of constant surface temperature of the cluster advanced at the beginning. Simply put, as soon as it gets warmer, the cluster has the "luxury" of warming up above this 7°C limit. From these levels, the equations should be taken up differently, but we also know that this case is less constraining for the bees.

In conclusion: when to insulate your hives?

Here is a nice model that allows to evaluate the effect of insulation on the dynamics of the bee colony.
We have learned and verified that:

  1. the roof of the hive is the main source of heat dissipation.
  2. the risk zone for bees starts for ambient temperatures below 5°C.
  3. insulation by a simple polystyrene panel brings a nice improvement of 50%.

Therefore in climatic zones where ambient temperatures fall below 5°C it is clearly useful to insulate your hives. For the more temperate zones the interest is less but certainly not negligible either. This would bring more comfort to the colony but maybe favors early queen rearing and varroa start. Which means there should be a choice to make.

Be careful to check the thermal resistance of your material (R = thickness / conductivity). Here are two examples:

  • Apifoam of 20mm : R = 0,020m/0,038W/m/K = 0,52
  • 40mm extruded polystyrene: R = 0.040m/0.033W/m/K = 1.20

Polystyrene, which is two times thicker, has more than double the thermal resistance because it also has a slightly lower conductivity.

Now that the insulation is in place, I hope that the bees will be able to get through this winter period better. It is -3°C tonight...

Extra note (winter 2023) : Since I installed this insulation in Jan 2021 it has been in place all year long. It is clearly improving the hive comfort and I am now feeling how hot is the inner side of the polystyrene at every inspection.


  • Camron James said:

    I don’t know if you’ll see this or not but I am curious how the insulation affected the interior moisture level in the hive. It looks like your design does not provide a space for ventilation between the hive and cover.

    Here in the states the general recommendation is to keep an upper entrance open all winter to improve ventilation, thereby reducing moisture buildup, especially on the inside walls and areas where water may drip onto the cluster.

    This advice has always seemed a little dubious considering most beekeepers slap a 10kg candy board on top of the whole stack that the bees are unlikely to even touch unless/until they use up their honey reserves. The sheer mass of sugar alone is going to absorb a ton of moisture over the winter and if you put a quilt box full of pine or cedar shavings on top of that you’re adding even more moisture absorbing capacity.

    March 02, 2024

  • Glise said:

    Bonjour, très interessant, est ce que par contre cela n’apporte pas de problèmes d’humiditée, car le polystyrene n’est pas un matériel respirant. Ne pourrait t’on pas remplacé ça par un matériel aussi isolant mais qui laisserait passer l’humiditée? (paille, copeaux, …?)

    March 02, 2024

  • Louis LOUPIAC said:

    Très belle étude, je partage totalement ce raisonnement. Si l’on compare avec ce qui se fait en matière d’isolation pour le bâtiment j’imagine une façon plus simple et vraisemblablement plus efficace encore en isolant par l’extérieur.
    J’ai réalisé une enveloppe avec du polyuréthane de 5 cm genre cloche fromage, facile à mettre et à enlever supprimant tous les ponts thermique.
    L’expérimentation ne m’a pas permis de tirer des enseignements car dès le printemps avant d’avoir pu mesurer le résultat, ne serait-ce qu’à l’œil nu, on avait retiré des cadres avec des abeilles pour renforcer d’autres essaims; Mais de ce fait j’en déduis que le résultat devait être positif sinon on n’aurait pas prélevé des abeilles.
    Je trouve votre étude très intéressante et je souhaite pouvoir continuer à échanger avec vous sur ce sujet.
    Je ne suis pas thermicien et je n’ai certainement pas vos compétences, je suis issu du bâtiment.
    Bien cordialement et à bientôt
    Louis LOUPIAC

    March 02, 2024

  • vialle said:

    vraiment tres interesante toutes ces observation et calcul cependant je constate que les parroies egalements de nos ruches ne sont pas suffisament isolante pour preuve on arrive a voir ruisseler de l eau ou moisir des cadre en hiver ce qui a mon avis n y est pas pour rien dans le developpement des locques en particulier dans certaines region plus froid.. merci encore pour la publication de cet article bien abouti

    March 02, 2024

Leave a comment