Go back to the home page for this issue of the Diplomate

Guest Column

Control Banding: Pharmaceutical Caterpillar to Mainstream IH Butterfly

By Ernest Sullivan, PhD, CIH, ROH, CChem and Om Malik, PhD, CIH, ROH, PEng

Control banding has experienced a favorable global reception. Three international workshops and the active consideration or promotion of control banding by organizations such as AIHA, ACGIH, IOHA, NIOSH, ILO, and WHO demonstrate that the concept merits the professional attention of occupational hygienists. One particular model has acquired a momentum of its own, to the extent that it has become the focus of almost all validation efforts by occupational hygienists. Other models are possible and we will present some considerations regarding a model with a local flavor.

Why Control Banding?

Compliance and management of risk from chemicals are of concern for occupational hygienists, employers and regulators alike. Control banding was developed to assist employers in small-to-medium enterprises to select appropriate control strategies. In this approach, control measures for chemicals are grouped into bands generated from a simplified risk assessment with parameters representing health hazard and exposure potential.

Control banding began in 1988 as a caterpillar in the garden of pharmaceutical big business. Emerging from its chrysalis in the late 90s, control banding came into the light of mainstream industrial hygiene, with the development of the COSHH Essentials model by the U.K. Health and Safety Executive (HSE).

The COSHH Essentials Model

The approach is to group chemicals into one of five health hazard bands (A to E) according to the increasing severity described in risk phrases applicable to the chemical. Risk phrases (or R-phrases) are a legislated requirement in Europe, assigned to products during classification by suppliers. Exposure potential is represented by considerations of quantity used and the qualitative degree of dustiness (if a solid) or volatility (if a liquid, with boiling point and operating temperature as surrogates), each generally with three levels. Combined in an empirically derived matrix, four control levels of increasing stringency are generated:

Level 1: General ventilation and basic industrial hygiene practice
Level 2: Local exhaust ventilation (engineering controls)
Level 3: Isolation/enclosure
Level 4: Determined by specialist advice     

Implicit in a control banding approach is the linkage of control measures to specific target ranges. Validation attempts generally support the HSE approach, not unexpectedly since the control level matrix array is based on professional experience, in order to produce a realistic outcome. Indeed, for any prospective model of control banding to attain general applicability, just such a pragmatic element must necessarily be incorporated.

Alternatives for Bands and Rankings 

Decisions have to be made regarding the number of bands and the number of levels in a particular band. But how many is enough? Philosophically, the number of assumptions required to describe any hypothesis should be minimized: Occam’s razor.

Distinct bands for quantity and dustiness/volatility are needed since these represent exposure potential. Regarding Health, COSHH Essentials was developed as an alternative to occupational exposure limits (OELs) and exposure assessments. Ultimately, however, compliance with the OELs must be addressed. Conceptually, OELs could be introduced as an explicit independent parameter, rather than as an implicit exposure target range; it is not “double dipping” since OELs (see later) are not synonymous with the highest health hazard classification.

How many band levels are necessary to conform to reality? COSHH Essentials has three for quantity, three for dustiness, three (plus temperature) for volatility, five for health and four for controls. However, the familiarity of three arbitrary levels – Low, Medium, High – is perhaps a normal human predilection.

Control

There are basically only three levels of control: Low involving no particular controls other than natural or dilution ventilation, Medium referring to local exhaust ventilation and High referring to enclosure or containment. “Specialist advice” is a fourth control strategy in COSHH Essentials. Realistically, however, design, implementation and verification of controls – particularly where there is almost certainly the potential for serious health consequences – are unlikely to be undertaken without professional advice.

Volatility

COSHH Essentials adds the additional factor of operating temperature; this refinement would affect some liquids with borderline volatility ratings. The model, consequently, must already be robust enough to accommodate minor deviations such that the same control measure will still apply to substances with similar volatilities but with arbitrarily different volatility ratings. (About 10 percent of liquids with listed TLV®s lie within a 10 C range, higher than the arbitrary division points of 50 and 150C.)

Health Hazard

A number of toxicity classification schemes incorporate five tiers. These can be combined into three broad bands, as shown below, since the distinctions between levels 1 and 2 (minimal/slight) and between levels 4 and 5 (serious/severe; highly/extremely) appear minor.

Table 1 – Toxicity Classification

North American workers are unfamiliar with R-phrases and the non-standardization of risk phrases on labels and material safety data sheets precludes precision regarding classification of health effects. The advent of the Globally Harmonized System of Classification and Labeling of Chemicals (GHS) may provide a regulatory stimulant to the use of standardized risk phrases in control banding, presupposing that the classification carried out by suppliers is accurate. 

Control Banding Model with a Maple Flavor
 
For any model, success or usefulness in implementation will ultimately depend on its conformity to professional experience, irrespective of the assumptions inherent in its derivation.

One assumption in this model is that the refinement of operating temperature is unnecessary and that boiling point provides an adequate basis for volatility rating.

In Canada, health classification under WHMIS is established and available. A three-tier rating system for health hazard based on WHMIS* is envisaged as:

• LOW  Not a WHMIS controlled product
• MEDIUM WHMIS classes D1B, D2B
• HIGH  WHMIS classes D1A, D2A, E

*D1A, D1B: acute toxicity; D2A, D2B: other toxic effects (mainly chronic); E: corrosive.
 WHMIS and the GHS have similar corresponding classification criteria.
 
Under this scheme, the distribution of the approximately seven hundred substances with listed Ontario OELs – essentially the same as TLV®s – is indicated below:

Table 2 – Health Hazard Rating Distribution

We have incorporated OELs into our model, for the reasons considered previously. Three levels are assigned, ranging from that representing the least harmful properties (Level A; high OEL) to the most harmful properties (Level C; low OEL):

C (LOW OEL)  Less than 50 ppm for gases and vapors
   Less than 0.5 mg/m3 for solids and mists
   Contaminants with Ceiling Exposure Values

B (MEDIUM OEL) Between 50 and 500 ppm for gases and vapors
   Between 0.5 and 5.0 mg/m3 for solids and mists

A (HIGH OEL)  Greater than 500 ppm for gases and vapors
    Greater than 5.0 mg/m3 for solids and mists
    No OEL listed and no available toxicological data

As shown in the following table, the distribution of OEL ratings is quite different from that based on health rating. In fact, for nearly half the listed substances, the corresponding ratings for Health and OEL are different. 

Table 3 – OEL Hazard Rating Distribution

Schematic for Risk Assessment and Control Selection



Example of Derivation of a Risk Matrix for Exposure Potential

Combined, the elements of dustiness/volatility and quantity represent an index of potential for exposure. This categorization is intuitive, based on assigning equal importance to quantity and dustiness/volatility. The combination of low dustiness/volatility with small quantity ranks lowest (1) and high dustiness/volatility with large quantity ranks highest (3). The combination of medium ratings for both dustiness/volatility and quantity is intermediate (2). The two combinations – high dustiness/volatility with small quantity, and low dustiness/volatility with large quantity – will have rankings equal to the latter. The remaining matrix elements have been apportioned so that there is uniform distribution – three matrix elements – over each of the three broad categories, shown as follows:

Table 4 – Risk Matrix for Exposure Potential
 

DUSTINESS/VOLATILITY	HIGH	2 High/Small	3 High/Medium	3 High/Large
	MEDIUM	1 Medium/Small	2 Medium/Medium	3 Medium/Large
	LOW	1 Low/Small	1 Low/Medium	2 Low/Large
		SMALL	MEDIUM	LARGE
		Quantity ®

The matrix for risk level is generated similarly.

The decision matrix for control selection is determined by combining the risk Level matrix with OEL rating; the distribution of elements in the matrix is empirical, reflecting field experience representing realistic controls.

Table 5 – Decision Matrix for Control Selection

Selection begins at the center, with the appropriate sector for quantity then proceeding outwards progressively selecting the appropriate sectors in the subsequent rings for dustiness/volatility, health hazard and OEL. The control selection appears in the outer ring.

Gases

About 10 percent of the substances with listed TLV®s are gases. Gases do not fit neatly into a control banding approach. About one-third would require isolation, if controls were determined by their intrinsic health (or OEL) ratings, but practical experience indicates that engineering controls would generally suffice. Gases should therefore be treated independently.

Web Application

We are developing a comprehensive web-based approach that includes health hazard classification, OELs, boiling points, CAS numbers and synonyms in recognition that most of such information is available in the literature and that from safety data sheets is often seriously deficient. The prototype web calculator can soon be accessed from the website http://www.ecoh.ca

Where to Now?

• A working group under AIHA is developing a framework document for control banding.
• The advantages and lessons learned from corporate control banding programs need to be consolidated and disseminated.
• International organizations will continue with development and implementation of Chemical Control Toolkits.
• A critical evaluation of models and examination of their internal validation are essential.
• Control banding is particularly well suited to a web-based application.

Sullivan is a consultant with E.A. Sullivan & Associates, London. He can be reached at esullivan@ody.ca.

Malik is principal and CEO of ECOH Management Inc. in Toronto. He can be reached at omalik@ecoh.ca.