Thermal insulation: Why is the R-value no longer enough?

Thermal resistance R (a key parameter in building codes for decades) measures a material’s ability to resist heat transfer by conduction under static, controlled laboratory conditions. It says nothing about what happens when the insulation is installed and exposed to temperature fluctuations, humidity, attic ventilation, or internal air movement. Yet it is precisely under these real-world conditions that the physical properties of materials diverge, sometimes dramatically.

This article draws on factual data and scientific studies to document this gap and identify the additional criteria (density, phase shift, humidity, effusivity, carbon footprint) that must now be taken into account when specifying high-performance insulation.

1. What recent studies reveal about the actual performance of insulation materials

1.1 A documented gap between theoretical gains and actual savings

In July 2025, the National Observatory for Energy Renovation (ONRE) published a study based on data from 80,000 single-family homes equipped with Linky and Gazpar smart meters. The results are significant: after insulation work supported by the CEE and MaPrimeRénov’ programs, actual energy savings represent only 36% of the theoretical gains for electricity and 47% for gas.¹ For every kilowatt-hour saved as reported by the standardized CEE data sheets, less than half is actually reflected on utility bills.

The authors identify several contributing factors: overly optimistic initial models, inconsistent workmanship, and—most significantly—the intrinsic properties of the materials, which do not behave under real-world conditions in the same way as they do under laboratory test conditions.

1.2 A persistent feeling of thermal discomfort

The second round of the ADEME “Sustainability and Lifestyles” survey (Summer 2025), conducted among 4,000 people representative of the French population, reveals that one in three French people consider their homes to be poorly insulated against both cold and heat.² This finding comes against a paradoxical backdrop: billions of euros have been invested in energy-efficient renovations over the past fifteen years, certifications have proliferated, and yet the perceived thermal quality remains inadequate for a large portion of the population.

1.3 Regulatory calculation methods under scrutiny

The Rivaton report, submitted to the Ministry of Housing in July 2025, points to the RE2020’s“still incomplete climate adaptation,” particularly regarding summer comfort. It notes that the expected theoretical carbon savings could be 15 to 25% lower in reality, due to discrepancies between the declared values of building components and their actual performance in use.³

In its 2025 white paper, the Association of Bio-based Construction Manufacturers (AICB) offers a complementary analysis: while the RE2020 calculation tools incorporate the concept of summer comfort, they do not take into account all the material characteristics that influence this phenomenon, particularly the thermal phase shift of walls and moisture-regulating capabilities.⁴ A shortcoming that the association calls for addressing.

The IFPEB and Carbone 4, through their Low-Carbon Prescribers Hub, also confirm that only 8 out of 25 projects analyzed already meet the RE2020 thresholds set for 2025–2028, and that the use of bio-based materials is the primary lever for action available.⁵

These publications all point to the same fundamental conclusion: current methods for evaluating insulation do not account for all the physical phenomena at work under actual conditions of use. To understand why, we need to examine exactly what happens in unheated attics.

2. Internal convection in unused attic spaces: what the Tampere study measures

2.1 Study Protocol and Context

In 2019–2020, Henna Kivioja and Juha Vinha, researchers at the Faculty of Building Physics at the University of Tampere (Finland), published a rigorous study on internal convection in highly insulated unheated attic spaces in the international scientific journal *Energy & Buildings* (Elsevier).⁶ This research is part of the European public project COMBI (Comprehensive development of nearly zero-energy municipal service buildings), funded by the European Regional Development Fund and the Finnish agency TEKES, with the participation of 37 companies . The authors have no declared conflicts of interest.

The protocol is rigorous: a 5 m² structure replicating an open-rafter floor (drywall, airtight membrane, blown-in insulation) is tested in a calibrated hot box (EN ISO 8990 method), with precise measurement of energy loss. Twenty-four configurations are tested, varying the following factors: the type of insulation (blown glass wool at 25 kg/m³ / cellulose insulation at 40 kg/m³), the thickness (300 and 600 mm), the temperature difference (20°C and 35°C), attic ventilation, and the presence or absence of roof trusses.

2.2 Results that call into question the regulatory calculation methods

The measurements yield results that call into question the reliability of the R-value as the sole criterion for performance in attic spaces:

• All configurations tested using blown glass wool result in energy overconsumption ofat least 10% compared to the theoretical calculation, due to convective movements within the insulation itself.
• In the worst-case scenario (ventilated attic, 600 mm thick, large temperature difference), actual energy consumption is up to 63% higher than predicted by the calculation based on the declared R-value.
• For a temperature difference of just 20°C (common in winter in France’s temperate regions), blown-in glass wool allows up to 46% more energy to escape than expected.
• This phenomenon is observed in 80% of the glass wool configurations studied, even when the parameters appear to comply with the thresholds set forth in the EN ISO 10456 standard, which directly calls into question the relevance of this regulatory standard.
• For cellulose insulation (density 40 kg/m³): this phenomenon is absent or negligible in virtually all scenarios (0 to 16% at most, under the most extreme conditions with a temperature difference of 35°C).

2.3 Scope and limitations of the study: as stated by the authors themselves

The study’s authors are clear about the scope of their findings: these “cannot be directly generalized to all blown-in loose-fill insulation, because even at the same density, the microscopic texture may differ.” They therefore limit their conclusions to the materials and installation methods specifically tested in the COMBI and FRAME projects. This caveat is essential to keep in mind when interpreting the data: the study raises a legitimate scientific question regarding the adequacy of calculation standards; it does not condemn a category of materials in absolute terms.

Furthermore, the COMBI study is part of a body of converging Nordic research. The work by Gullbrekken et al. (Norway, 2017 and 2019), published in the Journal of Building Physics and cited in the Tampere study, found that significant internal convection in 500 mm glass wool insulation occurs at a modified Rayleigh number as low as 4, a value well below the regulatory threshold of the EN ISO 10456 standard, set at 15.¹² These results reinforce the argument that current standards systematically underestimate the phenomenon. The same authors also observed a significantly weaker convective effect in cellulose insulation in mineral wool insulation, confirming the role of the material’s density and texture.

Finally, the FRAME study by Vinha et al. (University of Tampere, 2013)—a precursor to the COMBI project—had already demonstrated internal convection in both types of materials tested (glass wool and cellulose insulation) when the insulation was installed manually. It is the use of a blowing machine, specifically adapted for loose-fill insulation, that partly explains the more favorable results obtained forcellulose insulation the COMBI study: the texture of the insulation differs depending on the installation method.⁶ This point underscores that performance can never be reduced to the material alone: the quality and method of installation play a distinct role for all insulation materials.

2.4 The mechanism in question: internal convection

The physical phenomenon behind these temperature differences is known as internal convection. In an unheated attic during the winter, a temperature difference develops between the bottom of the insulation layer (the warm side, in contact with the living space) and the top (the cold side, in contact with the attic space). In a low-density, porous insulation material, warm, light air rises to the center of the layer, while cold, dense air sinks back down to the edges. These swirling movements within the insulation itself transfer heat much more efficiently than simple conduction. As a result, the effective thermal resistance is reduced, sometimes quite significantly.

This phenomenon is not detectable in laboratory tests, which measure thermal conductivity under static and controlled conditions. It is not captured by the declared R-value or by ACERMI certification, whose testing protocol does not replicate the dynamic conditions of an attic under real-world weather conditions. This is, in fact, the explicit conclusion reached by the authors of the Tampere study, who call for a “critical evaluation” of the EN ISO 6946 and EN ISO 10456 standards to determine whether they adequately account for the effects of internal convection.

cellulose insulation resistance cellulose insulation this phenomenon can be explained by its higher density and fine, fluffy texture, which mechanically impede internal air movement. The authors explicitly establish this link between density and convective behavior, while noting that the installation method also plays a specific role in the behavior ofcellulose insulation.

3. Additional criteria for evaluating the actual performance of insulation

The R-value remains a necessary indicator, but it is insufficient on its own to predict how insulation will perform under real-world conditions. Other parameters should be systematically included in the assessment.

Density: a key factor in convection and a prerequisite for the validity of the certified lambda value

This is the main conclusion of the Tampere study, which is independently confirmed by the work of Gullbrekken et al.: for blown insulation in unused attic spaces, density directly determines resistance to internal convection. Below a certain threshold, the declared R-value will not be achieved in actual conditions, regardless of the quality of installation.

Field data from France provides a concrete measure of this issue. The official technical documentation approved by the CSTB for blown glass wool in unused attic spaces indicates an in-place bulk density of 11 to 15 kg/m³.¹¹ The Tampere study tested this material at 25 kg/m³, which is approximately twice as dense.

The installation density also has a direct impact on the reliability of the ACERMI-certified thermal resistance: for loose-fill insulation, the declared λ value is certified at a specific nominal density. Installation at a density lower than this reference value (which can occur if the blowing machine is not properly adjusted) mechanically results in an actual thermal resistance lower than the calculated R-value, without this being attributable to the product itself. This is another reason why the actual installation density should be verified and documented on every job site.

cellulose insulation is ACERMI-certified for densities ranging from 30 to 40 kg/m³ depending on the thickness.¹³ It is within this density range that the Tampere study confirmed the absence of significant convective phenomena. With R-values declared equivalent, the in-situ thermal performance of the two categories of materials therefore diverges on this specific criterion, as documented.

Hygrothermal behavior: managing water vapor

Blown glass wool is, by nature, highly permeable to water vapor. The Technical Application Document approved by the CSTB explicitly states: “Once installed, the product is highly permeable to water vapor.”¹¹ It is precisely for this reason that DTU 45.11 requires the installation of a vapor barrier in many configurations of unused attic spaces. Any failure in the sealing or installation of this vapor barrier exposes the insulation to moisture absorption, which significantly degrades its thermal performance, as wet mineral wool loses a significant portion of its thermal resistance—and this loss is irreversible if drying is not possible.⁷

cellulose insulation fundamentally differently: as a hygroscopic material, it naturally absorbs and releases water vapor without any loss of performance or structural damage. This moisture-buffering capacity helps regulate indoor humidity and reduces the need for restrictive vapor barriers in compatible applications. The AICB emphasizes that this moisture-regulating property is not accounted for in current RE2020 calculation tools, leading to an underestimation of the actual performance of hygroscopic materials.⁴

Thermal lag and effusivity: criteria for summer comfort

Thermal lag measures the time it takes for a heat wave to pass through a wall. The higher the value, the later the daytime heat reaches the interior—ideally after the nighttime coolness has allowed the accumulated heat to dissipate.Effusivity measures a material’s ability to absorb and release heat to its surroundings. These two properties, which can be calculated according to ISO 13786, are not reflected in the R-value yet are crucial for summer comfort.

They depend directly on two physical parameters: the material’s density and its specific heat capacity. Regarding the latter, the available technical data is consistent: the specific heat capacity of cellulose insulation from 1,900 to 2,100 J/kg·K, which is roughly twice that of mineral wool (approximately 1,000 J/kg·K).⁸ Combined with its significantly higher installation density (30–40 kg/m³ versus 11–13 kg/m³ for blown glass wool), this property results in a very different thermal phase shift: at an equal thickness of 300 mm in unused attic spaces, blown glass wool provides a phase shift of about 3 hours, compared to about 9 hours for cellulose insulation .⁸ The AICB confirms that these phase shift properties are not properly incorporated into current RE2020 calculation tools, to the detriment of an accurate assessment of summer comfort.⁴

Life Cycle Carbon Footprint (FDES)

With the tightening of the RE2020 thresholds (the Ic construction indicator has dropped from 640 to 530 kg CO₂eq/m² for single-family homes as of January 1, 2025, and it will decrease further in 2028 and 2031⁹), the selection of insulation is no longer just a thermal equation. It is also a carbon equation.

cellulose insulation a twofold advantage here: very low embodied energy (made from recycled paper) and the ability to sequester biogenic carbon. For a 100 m² home insulated with approximately 1 ton of cellulose insulation, nearly 1,370 kg of CO₂ equivalent is stored in the walls throughout the building’s lifespan, a direct asset for meeting the RE2020 2025-2028 thresholds.¹⁰

4. Toward a Multi-Criteria Evaluation Framework: Practical Application

Factors to consider in any analysis of attic insulation performance

Based on the data presented in this article, a rigorous assessment of the performance of blown-in insulation in unused attic spaces should address the following questions:

1. What is the actual density of the blown-in insulation in the attic?
2. What is the thermal lag of the insulated wall?
3. Does the material manage water vapor without the need for a vapor barrier?
4. What is the product’s FDES carbon footprint in relation to the RE2020 construction thresholds for 2025–2028?

Notes and sources

1. INSEE / ONRE (National Observatory for Energy Renovation), July 2025. Effects of residential thermal insulation on actual residential energy consumption. Working Paper No. 2025-16. Study based on data from Linky and Gazpar smart meters, covering approximately 80,000 single-family homes between 2018 and 2023. insee.fr
2. ADEME / ObSoCo, November 2025. Barometer of Simplicity and Lifestyles — French People’s Practices, Perceptions, and Aspirations Regarding Simplicity. Survey conducted in the summer of 2025 among 4,000 people representative of the French population aged 18 to 75. librairie.ademe.fr
3. Robin Rivaton, July 2025. Report on the Evaluation of the 2020 Environmental Regulations. Study commissioned by Housing Minister Valérie Létard, submitted on July 10, 2025. 66 pages. ecologie.gouv.fr
4. AICB (Association of Bio-based Construction Manufacturers), 2025. White paper: “Summer Comfort: The Advantage of Bio-based Materials.” A 24-page document intended for project management and specification professionals. Reference: Olivier Joreau, President of the AICB. cahiers-techniques-batiment.fr
5. IFPEB / Carbone 4 — Low-Carbon Decision-Makers Hub, October 2024. RE2020 Feedback: Time for the First Carbon Footprint Assessment. Analysis of approximately 50 real-world projects in multi-family housing and the commercial sector. ifpeb.fr
6. Kivioja H. & Vinha J., 2020. Hot-box measurements to investigate the internal convection of highly insulated loose-fill roof structures. Energy & Buildings, vol. 216, 2020, 109934. University of Tampere, Finland. COMBI project funded by the European Regional Development Fund, TEKES, 9 cities, and 37 companies. No conflicts of interest declared. DOI: 10.1016/j.enbuild.2020.109934. French translation available from ECIMA (European Cellulose Insulation Manufacturer Association).
7. DTU 45.11 (NF, March 2020) and DTA Comblissimo (CSTB, Technical Approval 20/19-415_V1). The DTA specifies that blown glass wool is “highly permeable to water vapor” once installed, which requires a vapor barrier in the relevant configurations. In the event of wetting, the performance of mineral wool deteriorates irreversibly if drying conditions are not met, a principle documented in DTU 45.11 and by industry experts (source: Conseils-thermiques.org, based on best practices). cstb.fr
8. Specific heat capacity data and phase shift calculations according to ISO 13786. Specific heat capacity of cellulose insulation : 1,900 to 2,100 J/kg·K; mineral wool: approximately 1,000 J/kg·K (source: renovation-ecologique.fr, based on manufacturer data and ACERMI technical data sheets; values consistent with CSTB Technical Assessment Thermofloc 20/05-91, cited in forumconstruire.com). Thermal phase shift calculated at 300 mm thickness in unused attic space: blown glass wool (~11–13 kg/m³) ≈ 3 hours; cellulose insulation (~35 kg/m³) ≈ 9 hours. These orders of magnitude are consistent across multiple technical sources. Standardized calculation method: ISO 13786:2017.
9. Decree No. 2024-1258 of December 30, 2024, amending the energy and environmental performance requirements for building construction in metropolitan France. Effective January 1, 2025. Ic threshold for single-family homes: 640 → 530 kg CO₂eq/m² (−17%). Target threshold for 2031: 415 kg CO₂eq/m². ffbatiment.fr
10. FDES data on cellulose insulation, based on INIES data. Biogenic carbon storage calculation: for a 100 m² house insulated with 1 ton of cellulose insulation equivalent to approximately 900 kg of recycled paper), approximately 1,370 kg of CO₂eq is stored over the building’s lifetime. Source: Igloo France Cellulose, INIES FDES data, April 2024. cellulose-igloo.com
11. Saint-Gobain Isover / CSTB, Technical Application Document (DTA) for Comblissimo. Technical Assessment 20/19-415_V1, CSTB. Product specifications: “Apparent density in situ: approximately 11 to 15 kg/m³ (+5%)”. DTA available at cstb.fr
12. Gullbrekken L., Uvslokk S., Kvande T., Time B., 2017. Hot-box measurements of highly insulated wall, roof, and floor structures. Journal of Building Physics, 41(1), 58–77. And: Gullbrekken L., Grynning S., Gaarder J.E., 2019. Thermal performance of insulated constructions — experimental studies. Buildings, 9(49). Studies confirming significant natural convection in glass wool with a modified Rayleigh number as low as 4, and a lesser convective effect incellulose insulation. Cited in Kivioja & Vinha (2020).
13. ACERMI, reference table for blown insulation in unused attic spaces. For cellulose insulation , certified densities: 30 to 40 kg/m³. acermi.com

This article is based on verified scientific and institutional sources. The data from the study by Kivioja & Vinha (2020, *Energy & Buildings*, Elsevier) are taken from the French translation produced by ECIMA (European Cellulose Insulation Manufacturer Association). The results of this study are presented within the scope of validity defined by the authors: they apply to the materials and configurations tested and cannot be generalized to all blown-in insulation products. No manufacturer or competing product is mentioned or implicated in this article, which relies exclusively on generic technical categories.