Celotex Lambda Announcement
U-value CalculatorStart your U-value calculation >
Celotex announcement – Declared lambda value update
On 20 November 2017, we announced that we had identified a compliance issue with the declared lambda values of our 4000 range, 5000 range and Crown Bond and Crown Fix products. We confirmed that we had notified our regulator and the BBA. Following discussion with the BBA at that point we continued to market the 4000 range and Crown Bond and Crown Fix products with a lambda of 0.022 W/mK, pending further assessment. Since then, the BBA has audited us to assess the issue and define recommended improvements, which we have begun to implement. We have also done work to assess the impact on U-value calculations which we comment on further below.
The BBA certification of these products will continue on the following basis:
1. 4000 range and Crown-Bond and Crown Fix: all products manufactured in the XR4000 range, Crown-Bond and Crown-Fix ranges and GA4000 from 100mm and above will continue to be marketed as the 4000 range and Crown Range with a declared lambda of 0.022 W/mK.
2. New 3000 range: all other current products from our TB4000, GA4000, PL4000 and CW4000 ranges manufactured after 15 December 2017 will be marketed from January as Celotex 3000 with a declared lambda of 0.023 W/mK.
This adjustment arises from a change we are making (having gained more clarity on the relevant Standard) to our method for calculating lambda. It has a greater impact on products made on one of our production lines (typically our thinner products). We are continuing to work with the BBA with the aim of achieving lambda 0.022 W/mK for the products that will form our 3000 range in 2 above but we anticipate this could take some months.
The declared lambda is a component of any U-value calculation. Each U-value calculation is individual but, in almost all instances that we have looked at over the last weeks, the impact of the increase in declared lambda value has not been significant. We have been providing support to any customers who want to work through their U-value calculations.
Our focus is on ensuring that distributors, customers and end-users have continued access to certified products and all necessary product information so that they have full confidence in our range. We are updating our website to assist customers in relation to these matters, and we will continue to provide updates as appropriate.
Celotex announcement – Declared lambda value
20th November 2017
We have recently identified a compliance issue relating to our calculation and testing of the declared lambda value of products in the 4000 and 5000 ranges and the Crown-Bond and Crown-Fix products within the Crown Flat Roofing range.1 Based on the testing carried out, the effect of this is that for some years the declared lambda value in relation to the 4000 range has been understated by 0.001Watt/m/K 2 and in relation to the 5000 range has been understated by 0.001 – 0.002Watts/m/K3 4 . The declared lambda value of a product is a measure of the thermal conductivity of a sample as tested. The lower the declared lambda value, the greater the thermal efficiency of the product sample tested. We have notified our regulator and the BBA (our certifying body) of this and a copy of this announcement will be posted on the BBA website.
As we currently understand it, the issue relates to the completeness of the internal test data used for the declared lambda value and, for a period of time, the inclusion of certain test data which we now think should not have been used. The implications and effects of this are different for the 4000 and 5000 ranges and work to consider them is not yet complete.
We recognise that this is a serious matter which will prompt questions from customers. We are making this announcement so that our customers have relevant information about the declared lambda value not because of concerns about the safety of our products.
We believe that the issue affects products manufactured in the period between 2009 and October 2017 (it may extend earlier but we are still investigating this). We provide further information on the effect on the calculations in the tables below. The thermal efficiency of a building depends on a number of factors, of which the declared lambda value of the insulation material is only one.
We are working hard to complete our understanding of the issue and we continue to analyse the historical data to get the most complete picture we can. We are proposing to obtain independent assurance as to our processes. Our immediate priority is on products which we are selling now and we explain further below how this affects the 4000 range and Crown-Fix and Crown-Bond products. The suspension of the 5000 range will continue and now includes the FI5000 and GD5000 products.
4000 Range and Crown-Fix and Crown-Bond Products
The effect of the calculation issues we have identified is that the declared lambda value of the 4000 range and Crown-Fix and Crown-Bond products has been understated for some years (by 0.001Watt/m/K)5 6 if the full test data held by the business is considered. The lower the declared lambda value the greater the thermal efficiency of the product sample tested. The example and table in the Appendix explain the impact of the issue on the declared lambda value.
Since the identification of the issue, we have been doing further internal testing work which includes making some adjustments to the density of the product tested. Extrapolating from the current results of our internal testing, our indicative results suggest that the 4000 range and Crown-Fix and Crown-Bond products will meet its declared lambda value of 0.022Watts/m/K assuming further results continue to reflect the current trend.
On that basis and after discussion with the BBA, we are continuing to market the 4000 range and Crown-Fix and Crown-Bond products with its existing declared lambda value at this stage. We are in discussion with the BBA about our processes going forward.
5000 Range (suspended)
The effect of the calculation issues we have identified is that the declared lambda value of the 5000 range has been understated for some years (by 0.001 – 0.002Watts/m/K) 7 8 if the full test data held by the business is considered. The lower the declared lambda value the greater the thermal efficiency of the product sample. The example and table in the Appendix explains the impact of the issue on the declared lambda value. Work is ongoing to determine the long-term solution to this issue and the suspension will remain in place in the meantime.
Establishing best practice for the future
For the 4000 range and Crown-Fix and Crown-Bond products we are reviewing our systems and processes to ensure that we have addressed all such issues for the future.
At present, we are focused on completing our investigation and correcting the issues associated with our current production. We will update customers regarding when we expect to conclude this work, as well as any decision as to the 5000 range, when we have more information.
We recognise the seriousness of the issues and we are very sorry for the disruption that this may cause our customers. We would like to reassure our customers that we are working hard to resolve the issues as quickly as possible and that we do not consider they affect the safety of our products.
We recognise that this will prompt many questions. We are making this announcement so that our customers have relevant information about these matters. We would like to reassure them that we will do everything we can to answer these questions.
For questions or concerns regarding this matter, please contact our Technical Centre: (+44) 01473 820850
For all media enquiries please call: (+44) 07387 411972.
The examples in this Appendix recalculate declared lambda using all the available historic test data which should properly have been included. These examples have been calculated by Celotex and have not been audited by a third party body at this stage.
The information reflected in Table 1 (relating to the 4000 range) and Table 2 (relating to the 5000 range) has been recently identified, and represents our current best understanding of the situation.
The information may change as we do further work. In particular:
• These tables are provided for illustrative purposes.
• We continue to retrieve and review historical data and to test methodology to get the most complete picture we can.
• The “Measured lambda” has been calculated as at 31 December for each calendar year (2011 – 2016) and as at 31 August for 2017. Throughout each year the measured lambda varies. For some of the years shown, taking an earlier calculation date leads to a different lambda value than shown in the table.
• We are in communication with the manufacturers of the thermal testing machines to understand tolerances of different machines and how this impacts on lambda test results (positively or negatively).
Example: 4000 Series
The process for calculating the declared lambda value requires that once the value is exceeded, the declared lambda value is rounded up to the next whole number.
Therefore, for a supply of GA4050 in 2011, where the measured lambda of the product that was supplied was in fact 0.02206Watts/m/K, this would be rounded up to 0.023Watts/m/K under the calculation process. It was sold with a declared lambda of 0.022Watts/m/K.
Table 1 shows the measured lambda value for each year as at 31 December for each calendar year (2011 – 2016) and as at 31 August for 2017. The numbers in the third and fifth columns indicate the lambda value that should have been declared (depending on product thickness).
Example: 5000 Series
As noted above, the process for calculating the declared lambda value requires that once the value is exceeded, the declared lambda value is rounded up to the next whole number.
Therefore, for a supply of FR5050 in 2016, where the measured lambda of the product that was supplied was in fact 0.02218Watts/m/K this would be rounded up to 0.023Watts/m/K under the calculation process. It was sold with a declared lambda of 0.021Watts/m/K. Table 2 shows the measured lambda value for each year as at 31 December for each calendar year (2011 – 2016) and as at 31 August for 2017. The numbers in the third and fifth columns indicate the lambda value that should have been declared (depending on product thickness).
1. Referred to in this notice as the “Crown-Fix and Crown-Bond products”.
2. Revised on 21 November 2017 from “1Watt/m/K” to state the correct value of the understatement.
3. As at the dates and on the basis set out the Appendix.
4. Revised on 21 November 2017 from “1 – 2Watts/m/K” to state the correct value of the understatement.
5. As at the dates and on the basis set out in Table 1 in the Appendix.
6. Revised on 21 November 2017 from “1Watt/m/K” to state the correct value of the understatement.
7. As at the dates and on the basis set out in Table 2 in the Appendix.
8. Revised on 21 November 2017 from “1 – 2Watts/m/K” to state the correct value of the understatement.