Surveys, inspections and sampling

Every safe asbestos project begins with knowing where the asbestos is. Surveys and sampling turn suspicion into information: which materials contain asbestos, how much, what type, in what condition, and where. This lesson explains the different kinds of asbestos surveys, how homogeneous areas are defined, how many bulk samples are needed, how laboratories analyse them, and how to read and use a survey report. It is written for supervisors and workers who rely on surveys, not for the accredited or qualified people who carry them out.

Types of surveys

In Ontario, the owner's record under section 8 of O. Reg. 278/05 is in effect a management survey that must be kept current, and the report under section 10 is a pre-work survey of material that may be disturbed. BC requires a qualified person to prepare an inventory of asbestos-containing materials before work that may disturb them, and WorkSafeBC guidance describes the qualifications and survey methods it expects. Other provinces use similar approaches, often through codes of practice or guidelines.

A management survey is not enough for renovation or demolition. Management surveys often avoid destructive inspection, so asbestos hidden inside walls, under floors or in equipment may be missed. Before work that will disturb the building fabric, the survey must cover every area and material that the work will reach, including concealed spaces.

Homogeneous areas

Sampling is organized around homogeneous areas. A homogeneous area, or homogeneous material, is a material that is uniform in colour, texture, appearance and likely date of installation. Examples are all the 9-inch floor tiles of one pattern installed in one wing, all the pipe lagging on one system, or the textured ceiling throughout one apartment. Materials that look different, were installed at different times, or are in different additions of a building are separate homogeneous areas, even if they serve the same purpose.

Defining homogeneous areas well is a professional skill. If two different materials are lumped together, a sample from one may give a false answer for the other. If one material is split into many areas, sampling costs rise without adding information.

How many samples

Ontario sets minimum numbers of bulk samples in Table 1 of O. Reg. 278/05. The numbers depend on the type of material and, for surfacing material, the size of the homogeneous area.

If any one of the samples from a homogeneous material is found to be asbestos-containing material, the regulation allows the whole homogeneous material to be treated as asbestos-containing without analysing the remaining samples. This is sometimes called the positive stop rule. The reverse is not true: to conclude that a material is not asbestos-containing, all the required samples must be analysed and all must be below the 0.5 per cent threshold.

import math

# Minimum bulk samples for surfacing material (Ontario Table 1 style, metric)
def surfacing_samples_metric(area_m2):
    if area_m2 < 90:
        return 3
    if area_m2 < 450:
        return 5
    return 7

# Benchmark used for AHERA school inspections (imperial)
def surfacing_samples_imperial(area_ft2):
    if area_ft2 < 1000:
        return 3
    if area_ft2 <= 5000:
        return 5
    return 7

for m2 in (40, 89, 90, 300, 450, 1200):
    ft2 = m2 * 10.764
    print(f"{m2:5d} m2 ({ft2:7.0f} ft2): metric table {surfacing_samples_metric(m2)}, "
          f"imperial table {surfacing_samples_imperial(ft2)}")

# Why more samples help: chance that n random samples all miss asbestos
# when only a fraction p of the area actually contains it (illustrative)
for p in (0.5, 0.3):
    for n in (1, 3, 5, 7):
        miss = (1 - p) ** n
        print(f"area {p:.0%} contaminated, {n} samples: chance all miss = {miss:.1%}")

   40 m2 (    431 ft2): metric table 3, imperial table 3
   89 m2 (    958 ft2): metric table 3, imperial table 3
   90 m2 (    969 ft2): metric table 5, imperial table 3
  300 m2 (   3229 ft2): metric table 5, imperial table 5
  450 m2 (   4844 ft2): metric table 7, imperial table 5
 1200 m2 (  12917 ft2): metric table 7, imperial table 7
area 50% contaminated, 1 samples: chance all miss = 50.0%
area 50% contaminated, 3 samples: chance all miss = 12.5%
area 50% contaminated, 5 samples: chance all miss = 3.1%
area 50% contaminated, 7 samples: chance all miss = 0.8%
area 30% contaminated, 1 samples: chance all miss = 70.0%
area 30% contaminated, 3 samples: chance all miss = 34.3%
area 30% contaminated, 5 samples: chance all miss = 16.8%
area 30% contaminated, 7 samples: chance all miss = 8.2%

The two tables give similar but not identical results because their thresholds are set in different units. The second part of the script shows why the minimums exist. If a material is not uniform, so that only part of an apparently homogeneous area contains asbestos, a single sample can easily miss it. Three, five or seven samples taken from locations spread across the area greatly reduce that chance. The probabilities are illustrative; real materials vary in many ways, which is why the defined minimums must be followed.

Taking bulk samples safely

Bulk sampling disturbs the material, so it must be done carefully by a trained person. Good practice includes:

• Planning sample locations to represent the whole homogeneous area, including different floors, rooms and runs.

• Keeping other people out of the immediate area during sampling and wearing appropriate respiratory protection and disposable coveralls.

• Lightly wetting the material with a fine mist and placing a small bag or container under the sampling point.

• Taking a core through all layers of the material, because asbestos may be in only one layer.

• Sealing the sample in a labelled, airtight container, and sealing the sampled spot with a patch, tape or sealant.

• Cleaning tools and the area with damp wipes or a HEPA vacuum, and recording the location of each sample on a plan.

Workers who are not trained to sample should never take samples themselves. If a suspect material is found during work, stop and call the person responsible for asbestos assessment.

Laboratory analysis

Ontario's regulation requires bulk samples to be analysed by the method in US EPA Test Method EPA/600/R-93/116, which uses polarized light microscopy (PLM). Other provinces use the same or equivalent methods. Laboratories should be accredited for bulk asbestos analysis; many labs serving Canadian clients hold accreditation under the US National Voluntary Laboratory Accreditation Program (NVLAP) or an equivalent scheme.

Non-friable organically bound materials, such as vinyl floor tiles and mastics, can give false negatives by PLM because the asbestos fibres are very fine and coated by the binder. A negative PLM result for a floor tile or mastic should be treated with caution; many consultants confirm such results by TEM or gravimetric methods before declaring the material asbestos-free.

Reading a survey report

A good survey report gives workers and supervisors what they need to plan safely. It should include the building and areas surveyed, the date, the name and qualifications of the surveyor, the methods used, any areas not accessed, a list of homogeneous materials with sample results, the asbestos type and percentage, quantity, condition and location of each asbestos-containing material, drawings or photos, and recommendations.

Supervisors should check four things before relying on a survey:

• Scope. Does it cover all the areas and materials the work will disturb? A management survey of accessible areas is not a pre-demolition survey.

• Limitations. Which areas were not accessed, such as wall cavities, locked rooms, live equipment or roofs? These must be treated as unknown.

• Date. Has anything changed since the survey, such as earlier renovations, damage or additions?

• Results. Which materials are asbestos-containing, which are assumed, and which tested negative? Assumed materials must be handled as asbestos-containing.

Managing asbestos left in place

Many owners choose to leave asbestos-containing material in place when it is in good condition. This is often the safest and least costly choice, but it creates a long-term duty to manage the material. A management program, sometimes called an operations and maintenance program, keeps the survey up to date, labels or marks known materials where practical, inspects their condition on a schedule, controls work that could disturb them through work permits, trains maintenance and custodial staff, and specifies how small repairs and clean-ups are done. Every contractor who works in the building should receive the relevant survey information before starting, so that the material left in place does not become the next accidental disturbance.