6. Chemical Analyses

Before collecting building material or air samples, a testing laboratory should be selected. This section details how to ensure that sampling is effective, and the exercise produces useful test results. Surveyors should be aware of the role of the testing laboratory and the content of the chemical analyses.

The collection of air samples is described in Section 7, Determining PCB Content in Indoor Air, while the collection of samples from building materials is discussed in Section 8, Determining PCB Content in Building Materials.

Analysis laboratories handle samples differently and this may have a bearing on how much testing material is required. When selecting an analysis laboratory, please note the following:

Some analysis laboratories are accredited, meaning that a national accreditation body has approved the competence and impartiality of the laboratory. In Denmark, the Danish Accreditation Fund (DANAK) accredits laboratories to conduct sampling and/or technical or chemical analysis of specific materials and products according to the accreditation standard DS/EN ISO/IEC 17025 (Danish Standards, 2005). The accreditation sets out the relevant PCB analytical chemical methods (e.g., PCB analysis of air samples). If an accredited analysis is required, the testing laboratory should be notified about exactly which sample type to analyse.

At minimum, the testing laboratory should have documented experience in analysing for PCBs in the specific sampling material in question (such as caulk or a sampling medium for air samples). The experience in some testing laboratories is restricted to PCB testing of foodstuffs, soil, and water.

According to Vejledning for måling af PCB i indeklimaet 2014 (Guidelines to Measuring PCBs in Indoor Climate 2014) (Danish Transport, Construction, and Housing Authority, 2015), sampling and chemical analysis should be carried out by a measuring firm and laboratory that maintains a quality assurance system according to DS/EN ISO/IEC 17025 (Danish Standards, 2005) with experience in measuring PCBs in indoor air.

Some testing laboratories are willing to help with advice and guidance on sampling. However, it is important to ascertain whether their prescriptions are based on standard methods, experience, or assumptions. If you collect the samples yourself, the responsibility for correct sampling rests with you.

To analyse air or building material samples, PCBs will first have to be extracted from the sample. Standards and guidelines are at hand and the laboratory method should meet the same degree of extraction efficiency as the relevant standard. DS/EN 15308 (Danish Standards, 2008b) is used for characterising PCBs in waste. The German standard VDI 2464 (Verein Deutscher Ingenieure, 2009) sets out how PCBs in air samples are extracted. The same applies to the American compendium EPA/625/R-96/010b (US EPA, 1999). 

When collecting air samples, there is a filter which collects airborne particles, and one or several adsorption media which collect gaseous-phase PCBs. It is important to ensure that the testing laboratory will analyse both the filter and the sampling medium.

A laboratory often offers various analytical testing packages and will often analyse for six or seven PCB congeners. PCBsum6 indicates the sum of the content of the six congeners PCB-28, PCB-52, PCB-101, PCB-138, PCB-153, and PCB-180. PCBsum7 also includes the congener PCB-118, which belongs to the group of dioxin-like PCB congeners. In the low-chlorinated PCB mixtures, there is, in practice, not much difference between PCBsum6 and PCBsum7, since the concentration of PCB-118 is often very low in relation to the dominant congeners.

The standard DS/EN 15308 for characterising PCBs in waste operates with PCB sum7 (Danish Standards, 2008b). Vejledning for måling af PCB i indeklimaet 2014 (Guidelines to Measuring PCBs in Indoor Climate 2014) (Danish Transport, Construction, and Housing Authority, 2015) refers to the procedure outlined in the German standard VDI 2464 (Verein Deutscher Ingenieure, 2009), which recommends that PCB-118 be included and PCBsum7 be determined.

The result of PCBsum6 or PCBsum7 only constitute part of the PCBs present in the product (see Section 1.1, Commercial PCB Products). If the sample is an air sample, PCBsum7 is multiplied by factor 5 to determine the total amount of PCBs (referred to here as “PCB-total”). For material samples, the testing laboratory will typically compare the analysis of the sample with different commercial PCB-products and is thus able to determine whether it is a known product. If a caulk sample matches one of the known products, one can know the total amount of PCBs in the caulk which comprises PCBsum6 or PCBsum7. Therefore, the result can be adjusted to reach an estimate of the total amount. If the sample does not match a known product, PCBsum6 or PCBsum7 will be multiplied by factor 5 to estimate the total amount (see Section 1.1, Commercial PCB Products).

A material analysis of PCBsum7 multiplied by factor 5 complies with the guidelines issued by the Danish EPA for investigation of PCBs (Danish EPA, 2011). If the analysis can identify the original product type, it is possible to apply the product type correction factor (see Section 1.1.5, Correction Factors). This may be relevant if the PCB concentration in the material is found to be close to an administrative concentration limit (e.g., of 50 mg/kg for hazardous waste) and a more accurate calculation is therefore needed (Danish EPA, 2011) (see SBi Guidelines 242, Renovering af bygninger med PCB, 4.2.1, Bestemmelse af koncentration (Renovating Buildings Containing PCBs, 4.2.1, Calculating Concentrations)) (Andersen, 2013b).

It is necessary to be clear about how small amounts of the substance should be traceable or detectable. The limit of detection is the lowest concentration determinable by the chemical analysis. Often, a limit of detection is requested that is ten times lower than the one which it is strictly necessary to detect. For example, this means that if one wishes to investigate whether waste can be designated as uncontaminated (i.e., with less than 0.1 mg PCB/kg waste (City of Copenhagen, 2014)), a limit of detection of 0.01 mg PCB/kg would be required. It makes a difference whether we are dealing with air samples or material samples. The recommended action value for indoor air issued by the Danish Health Authority is 300 ng PCB per m³ of air (Danish Health Authority, 2013a) and if 1 m³ of air is collected, this will typically require a limit of 30 ng of the total amount of PCBs to be detected per sample.

The limit of detection for PCB congeners will depend on the specific material type for which the congeners are determined (e.g., caulk, paint, or concrete).

Two methods are used for analysing PCBs:

GC-ECD is often cheaper than GC-MS whereas GC-MS is more specific. Generally, GC-MS is recommended, but there may be special conditions that favour GC-ECD (see Section 6.4, Sample Labelling).

For caulk, 5 cm (plus its entire depth) will normally suffice while concrete samples should weigh min. 10 g. Agreement should be made with the laboratory as to how much testing material is needed and the specific limit of detection required.

If the chemical analysis is performed in accordance with the standard DS/EN 15308 (Danish Standards, 2008b) on characterising PCBs in waste, concentrations as low as (or below) 0.01 mg/kg (dry matter) can be detected for each congener.

Regarding material recovery, concrete and brick will usually comprise most of the material recovered and the analytical limit of detection for these materials is therefore of interest.

If material samples are analysed to estimate whether the waste is recoverable (i.e., whether it contains less than 0.1 mg PCB per kg waste) (City of Copenhagen, 2014), it can be problematic if some of the seven indicator congeners are below the limit of detection, as this implies that the total amount of PCBs will remain unknown. When determining PCBsum7, the methods used to quantify congeners below the limit of detection are crucial. It is possible that the concentration of the congener which is below the limit of detection is set to zero (i.e., that it is not counted at all). It is also possible to configure the concentration as the limit of detection or perhaps half of the limit of detection. In this way, the congeners will be counted in when determining the PCBsum7 concentration.

When analysing air samples, a limit of detection of 30 ng/m³ for PCB-total (i.e., 6 ng/m³ for PCBsum7 or approx. 1 ng/m³ for the individual congeners) will often be sufficient (Verein Deutscher Ingenieure, 2009; Danish Transport, Construction, and Housing Authority, 2015). The limit of detection is reported as ng/m³ (i.e., allowances have been made for the test volume). Therefore, if 1 m³ is collected, it must be ensured that the testing laboratory operates with a limit of detection for individual indicator congeners of 1 ng/test tube.

The labelling of samples must be unambiguous, and we recommend that the testing laboratory are asked to state the location in the test report.

Most laboratories want to know whether there are special conditions with regards to the sample(e.g., for them to account for interference in the analysis). This could be data about paint on caulk or whether the sample has been in contact with oil. DS/EN 15308 (Danish Standards, 2008b) generally recommends the GC-MS method for material samples, but if the sample contains significant amounts of mineral oils, the standard recommends GC-ECD. The test report should state whether the sample analysis indicates interference and whether they have suspected reasons for this.

Material samples suspected of containing PCBs may turn out not to contain PCBs, but instead chlorinated paraffins. Chlorinated paraffins comprise a group of chlorinated compounds with a chain length between 10–30 carbon atoms and a chlorine content of 20–70 %. Chlorinated paraffins are used as replacement for PCBs and used in anti-corrosive paints and primers, plasticisers, and fire-retardants in plastics and caulk (Trap et al., 2006). The test report should always disclose this.

When taking aliquots of solid materials and pooling them into a composite sample (see Section 5, Mapping the Building Materials) the laboratory should be notified of this to ensure that the whole composite sample is represented in the material that is analysed.

Results may be reported in different ways in the test report. The PCB content is often stated for each of the six or seven congeners analysed for and the sum (PCBsum6 or PCBsum7). Alternatively, the total PCB content (PCB-total) and the factor for converting from PCBsum6 or PCBsum7 to PCB-total could be stated. For material samples, the factor is either determined based on the analysis of the composition of PCB congeners or set to 5 if the composition of the sample does not match a commercial product. For air samples, factor 5 is used.

Results of material samples can be reported in mg/kg or ppm (parts per million, which is no. of units per 1 million units). For concrete, the test result may be reported in mg/kg DM (dry matter).

1 mg/kg = 0.001 g/1,000 g = 1 g/1,000,000 g = 1 g/1·10⁶ g = 1 ppm. If the concentration equals 1 ppm, PCB comprise 1 g in 1 ton of material.

A PCB content (PCB-total) of 100,000 mg/kg (e.g., in caulk) corresponds to 100 g/kg (i.e., that 10 % of the weight of the caulk comprises PCBs).

Air-sample test results are reported in ng/test tube or ng/m³ if the total air volume was disclosed when submitting the sample.