Priority-setting regarding offshore substances and preparations

5. Conclusions

5.1 Pre-screening procedures

The pre-screening procedures developed in the present project are based on an evaluation of the intrinsic environmental and occupational health hazards of the individual substances present in the preparations and on the placement of these substances on lists of hazardous substances that are subject to phasing-out or cleaner technology considerations. The pre-screening procedure with regard to the environmental hazards does not take into account the amount of the substances used, the relative content in the preparation or the preparation itself whereas the pre-screening procedure with regard to the occupational health hazards is based on the classification of the preparation, in which information on both the human health classification of the individual substances and on their content in the preparation is taken into consideration.

In the pre-screening, offshore preparations are grouped into one of four groups. For environmental hazards: P1(E), P2(E), P3(E) or WL(E) with P1(E) being the group of preparations containing the most hazardous substances, P2(E) the group of preparations with less hazardous properties, P3(E) the group of preparations considered to be of low risk to the aquatic environment, and WL(E) a waiting list of preparations that cannot be grouped due to lack of data on the hazard of at least some of the substances. Similarly, a grouping is made for preparations according to occupational health hazards: P1(H), P2(H) and P3(H). Finally, the various preparations are placed in groups taking into account both environmental and occupational hazards, P1, P2, P3 and WL. Additionally, a preliminary group P2* is made for preparations that can be evaluated only with regard to health hazards (P2(H)) but not environmental hazards (WL(E)). Further data on environmental properties may place the preparations in either P1 or P2. It is the intention that offshore preparations grouped in P1 should be considered for cleaner technology measures or phasing-out and that preparations grouped in P2 should be further evaluated for their hazard or risk to the aquatic environment, e.g. according to the principles laid down in the CHARM model, or to health in the working environment.

273 offshore preparations containing 306 different substances were evaluated by use of the pre-screening criteria. Although the environmental hazard of 40% of the substances contained in the preparations could not be assessed due to lack of data, 10% of the preparations were grouped in P1(E), 8% in P2(E), 56% in P3(E), and only 26% could not be grouped due to lack of data on the environmental hazard of the substances. However, a more thorough data collection might have revealed more data resulting in fewer preparations on the waiting list.

Similarly, a grouping based on human health classification was performed resulting in 4% of the preparations placed in P1(H), 51% in P2(H) and 45% in P3(H). By combining the two criteria, 12% were grouped in P1, 35% in P2 and 28% in P3. Preliminarily, 12% were placed in P2* based on human health hazards but lacking data on environmental properties, and 12% were retained on a waiting list.

The use of the pre-screening procedures on 273 preparations has clearly demonstrated the applicability of the procedures for identifying the most hazardous offshore chemicals based on their intrinsic properties. The pre-screening requires that a data set is available corresponding to the HOCNF supplemented with a minimal data search for aquatic toxicity data of substances in the preparations.

5.2 Possibilities of substitution

Basic criteria for evaluation of possibilities for substitution were elaborated. The general rule is that P3 preparations are of low hazard and that they therefore whenever technically possible may replace P1 and P2 preparations. Similarly, P2 preparations may replace P1 preparations although differences in the used amounts should also be taken into account. Finally, specific criteria for comparison of two P1 preparations have been developed.

Preparations with identical main technical functions were compared according to their placing in the 3 environmental hazard groups. For some technical functions, both hazardous preparations and non-hazardous preparations are used in the offshore industry. Although no detailed comparison of the technical possibilities for substitution was possible, there might be some possibilities of substitution and these should be further evaluated. This is especially the case for preparations used as cementing chemicals and cement additives, gelling agents, viscosifiers and filtrate reducers, water-based muds and additives, weighting agents, and well stimulation fluids.

5.3 Comparison of PROSPECT and CHARM

The result of the PROSPECT environmental pre-screening of the preparations was compared with a generic risk assessment (i.e. environmental hazard assessment) following the principles of the CHARM model. Two comparisons were made:

1. A comparison of preparations with identical technical function (gelling agents and well stimulation fluids were selected)
2. A comparison of preparations that have already been replaced by other anticipated less environmentally hazardous preparations (7 pairs of preparations were selected).

In the latter comparison also the estimated sums of toxic units and bioaccumulative units were included in the evaluation.

Lack of data

Some general conclusions could be made from this comparison. In general, both the PROSPECT and the CHARM evaluations were hampered by lack of data on the ecotoxicological properties of both the preparations and of the substances contained in the preparations. The intention of the CHARM model is to use toxicity data for the preparations for predicting no-effect-concentrations for the various environmental compartments. Despite the criticism of this approach for not being scientifically valid, the use was evaluated but due to lack of those toxicity data, it was only possible to perform this assessment for very few of the preparations.

In stead, the evaluation was based on the most critical substance in the preparations, i.e. a combination of the intrinsic hazard of the substance, the content in the preparation and the estimated release of the substance. As no information on the recommended dosage was available, the so-called "critical dosage" was estimated for the preparations based on the content of the most critical substance. The comparison showed that even within the groups (P1(E), P2(E), P3(E)) very large differences in critical dosages were found.

The model requires aquatic toxicity data on either the preparations itself (although this approach as mentioned has some scientific limitations) or at least on the most hazardous substances contained in the preparations. However, for many of the preparations, information on the preparation or on the most critical substance was not available. Consequently, the estimated critical dosages of the preparations were not always based on the most critical substances and the critical dosages were therefore not directly comparable between the preparations.

Moreover, the CHARM model is only based on data on the immediate aquatic toxicity of the preparations/substances. This means that additional information on the environmental hazards of the chemicals is not taken into account. This is the case for alkylphenol ethoxylates, which are not as acutely toxic as some of the other substances used but which due to their potential long-term estrogenic effects are placed on the Esbjerg Declaration list of substances that should be phased out of actual use. This is also the case for some of the substances placed on the proposed EU list I. Some of these substances are not very toxic (and their placing on the list I might therefore be questioned) but as they are presently on the list, they are subject to cleaner technology considerations.

The present comparison of the PROSPECT pre-screening and the CHARM hazard assessment thus demonstrates that the pre-screening regarding environmental hazards was possible for the majority (74%) of the offshore chemicals despite the lack of data on 40% of the substances contained in the preparations while a precise CHARM hazard assessment could not be performed due to the lack of data. The CHARM model must therefore be considered more sensitive to lack of data than the PROSPECT environmental pre-screening model. This highlights the intention of the PROSPECT model as a pre-screening tool to be used for grouping of offshore chemicals in a group that should be considered for substitution (P1(E)), a group that can be used without further assessment (P3(E)), and a group that should be further evaluated (P2(E)). For the latter group, a more thorough evaluation could be made, for instance by use of the CHARM model.

Finally, it can be concluded that the PROSPECT and CHARM evaluation models are applying two different approaches. The PROSPECT approach (both in pre-screening and content of toxic and bioaccumulative substances) is focusing on the intrinsic hazard of the substances contained in the preparations, while the CHARM approach is focusing on a generic risk assessment of the use at a platform in the North Sea.

The two different approaches give different priority to some of the intrinsic properties of the substances. A high bioaccumulation potential, as determined by a high log P_ow, is negative in the PROSPECT approach, but in the CHARM evaluation (of production chemicals) these substances are estimated to be retained in the oil phase and they are therefore not considered to be of high risk for the aquatic environment. Similarly, substances of a low biodegradability are negative in the PROSPECT approach, but this is only considered in one compartment in the CHARM model, namely the hazard to the sediment compartment. Such substances might, however, cause a long-term risk for the aquatic environment.

Which of the two approaches that is most appropriate may of course be debated, but if priority is given to the phasing-out of the substances with the most hazardous environmental properties, the use of the PROSPECT approach for identifying such preparations seems more appropriate.