Incineration rules explained

The Waste Incineration and Large Combustion Plant directives are likely to impact on many operators, but what are the Environment Agency's requirements and how can mistakes be avoided?

Published 12th May 2006

Save

Register for free and continue reading

Join our growing army of changemakers and get unlimited access to our premium content

Login Register

The Dest Testing Association spells things out for edie.

The Waste Incineration Directive (WID) came into force on 28 December, 2005, while the Large Combustion Plant Directive (LCPD) will apply fully within another two years.

All of the installations covered by the WID, and LCPD installations above a certain thermal rating, will need to have continuous emission monitoring systems (CEMs).

These CEMs will in turn require quality assurance to EN14181, Quality Assurance of Automated Measuring Systems. So, during the next two years, a large and growing number of operators will have to apply EN 14181.

While there are now many test laboratories offering services to EN 14181, the Environment Agency (EA) and STA have observed a wide spectrum of quality in the application of the standard.

EN14181 specifies three levels of quality assurance, plus an annual surveillance test (AST). Under EN 14181, test laboratories have a critical role in the application of QAL2 and the AST for the calibration and verification of CEMs.

The core of both the QAL2 and AST procedures is a set of parallel tests using standard reference methods (SRMs). The purpose of the SRMs is two-fold.

Firstly, test laboratories use SRMs to establish the relationship – known as the calibration function – between the true value of the determinand in the stack, and the output of the CEM.

Secondly, the data from the SRMs can be used to determine whether the uncertainty of the CEMs is low enough, by analysing the scatter on the graph of corresponding measurements from the SRMs and CEMs.

This is known as the variability test. Valid calibration-functions and accurate variability-tests depend on a wide-enough spread of data, and an accurate and precise application of the SRMs.

The EA requires test laboratories performing services to EN 14181 to be accredited to ISO 17025, the MCERTS performance standards for manual stack monitoring, and the appropriate SRMs.

However, we have observed significant mistakes by some accredited laboratories, meaning that the application of QAL2 tests did not meet the requirements of EN 14181.

Mistakes made For example, even though EN 14181 requires a spread of data for a valid calibration function, this is often difficult to achieve in practice as a lot of data is seen in clusters.

Following trials of EN 14181, the EA, STA, process operators and test laboratories involved in these trials found the following:

  • If the data is clustered near zero, the calibration function of EN 14181 cannot be used and the test laboratory needs to perform a linearity test as a back-up for an approximate calibration

  • If the data is in a cluster higher up the scale of emissions, three values near or at zero are required so that the calibration function is not skewed

    Despite specifying these requirements in the guidance on EN 14181, the following examples have been seen several times:

  • Producing a calibration function by linear regression, even when the emissions are clearly too low or, worse still, actually zero

  • Not having values at or near zero when there is a higher cluster of data

  • Not examining the data to remove obvious outliers. For example, one test laboratory did not audit the CEMs data and did not remove hourly averages when the CEMs were carrying out an auto-zero procedure

  • Not performing a linearity test, even though historical site-data predicted low values for the emissions

    Even if the SRMs were performed to the required accuracy and precision in the above cases, the QAL2 and AST applications would be flawed. But there have been cases where even the application of the SRMs was questionable.

    For example:

  • During an audit of an installation, we found that in the QAL2 report, the test laboratory had mixed up the data from two stacks

  • When performing SRMs on an incinerator, the test laboratory’s reference instruments did not meet the MCERTS requirements for certification ranges. This meant that the reference instruments did not meet the WID requirements for uncertainty budgets.

    Topics

    Tags

    • Action inspires action. Stay ahead of the curve with sustainability and energy newsletters from edie

    Subscribe

    empowering sustainable business