Scaling Down the Pandacan Oil Depot in Manila

1. Background

After the terrorist attacks in the USA on 11 September 2001, the Government in the Philippines reviewed safety and security in the capital city, Manila. Because of its size, location and strategic importance to the fuel supply of Manila, the oil depot at Pandacan was identified as a potential security concern.

As a result, the three oil companies who occupy a 30-hectare industrial site at Pandacan (Pilipinas Shell, Caltex and Petron) agreed a strategy with the Mayor of Manila to scale down the site and create a green ‘buffer zone’ between the depot and the local community. Under the plans, the green buffer zone with a width of up to 50 metres would become a park for the use of the local community.

The creation of a community park from industrial land requires great care and attention. After more than 90 years of operation, there was little doubt that the industrial land would have some product residues that would need to be addressed before it could be declared safe for recreational use.

Pilipinas Shell commissioned international environmental consultants ENSR to carry out an environmental site assessment and a health risk assessment for the portion of Shell-operated land that was to be converted into a community park. The assessment led to a remediation strategy, which was carried out to remove these residues and improve soil quality before the park was landscaped and, ultimately, opened for public use.

Pilipinas Shell likewise decided to ask ENSR to conduct an off-site health risk assessment. This assessment involved sampling the soil and groundwater in Shell’s neighbouring communities, Barangay 833 and 834, for these residues that may have ‘migrated’ from the Shell site and conducting an analysis of what, if any, risk the presence may pose to the health of local residents.

2. What is health risk assessment?

The purpose of a health risk assessment of this nature is to analyse the health risk that is associated with any actual or potential -presence of product residues, based on test data and detailed ‘local knowledge’ such as how the land or groundwater is used, who used it and how often.

Risk assessments are conducted in various phases. The first phase is to analyse the levels of different substances found in soil and water samples against generic – and very conservative – risk levels. Many substances can, at this early stage, be eliminated from the risk assessment because levels found are so small that they pose no concern. Those remaining substances are then taken to a next stage of analysis, which is more detailed and more specific to the local area. This second stage involves gaining an understanding of all of the following variables:

• the nature (toxicity) of the substances ‘of concern’;
• the levels and locations of these substances found in tests;
• whether ‘pathways’ exist which might expose people to these substances (pathways might be, for example, inhaling dust, drinking water or eating livestock reared in the community);
• the profile of the local community and its inhabitants; and
• the habits of the community which could bring certain community groups into contact with the substances.

Once this information has been ascertained, the possible health risk can be reassessed. If a health risk is identified, action may need to be taken to mitigate the risk. This action might involve removing some of the substances and/or eliminating a ‘pathway’ (such as concreting exposed soil or restricting use of groundwater).

3. Appointment of an expert external review body

To provide an external perspective on ENSR’s health risk assessment methodology, analysis and conclusions, Shell asked the National Institutes of Health at the University of the Philippines to assemble an independent and expert review panel. 

The Health Technical Panel that was assembled – the first of its kind in the Philippines – included an Environmental Health Risk Assessment Specialist Prof. Ronald Subida, Environmental Health Specialist Engr. Elma Torres, Biostatistician Dr. Jesus N. Sarol, Health Policy Analyst Dr. F. Marilyn Lorenzo, and Toxicology Specialist Dr. Kenneth Hartigan-Go.  Technical Representatives from the Department of Environment and Natural Resources (DENR) and the Department of Health (DOH) were also present during the entire process as observers in behalf of their respective government agencies. 

Its purpose was to assess the technical soundness and validity of the on-site and off-site health risk assessments carried out by ENSR. This external review function would ensure that the technical work and risk assessment were carried out to a high standard.

A series of validation workshops were held with the Panel and ENSR – the third party consultant who conducted the Environmental Site Assessment and Human Health Risk Assessment (HHRA).

• Workshop 1: Defining the Risk Based Approach for the panel of experts (February 2003)
• Workshop 2: Discussion on the methodology for the conduct of an HHRA (April 2003)
• Workshop 3: Review of the On-site Health Risk Assessment (July 2003)
• Workshop 4: Review of the Off-site Health Risk Assessment (September 2003)
• Workshop 5: Review of the Post Remediation Health Risk Assessment (18 May 2004)

On May 2003, the Panel visited Pandacan to see the ongoing construction of the Linear Park Phase One and the two communities.  They also gained first hand experience on the health and safety practices onsite, updated remediation works and ongoing ESA activities.

4. The Health Panel’s review and input

The first draft of the off-site Heath Risk Assessment (HRA) for Pandacan was presented to the Panel on 9 October 2003. The conclusion of the HRA was that, while tests showed that some traces of residues on-site were found to have migrated offsite, they were at low levels that do not present a risk given the facts established about the local community’s profile and habits. 

The draft HRA was conducted to be consistent with international guidelines such as those from the US Environmental Protection Agency (EPA), the Netherlands, the UK and the World Health Organisation. The Panel conducted a review of related literature to assess the scientific soundness of the methodology used and the assumptions made in the study.

The HRA was subdivided into four main categories, which followed the internationally accepted four-step process for health risk assessments as developed by the EPA. The Panel’s overall judgement of the HRA would depend on how well these various components worked together to justify the conclusion drawn. The categories are:

• data evaluation and hazard identification;
• exposure assessment;
• toxicity assessment; and
• risk characterisation.

Sections 4.i to 4.iv summarise the Panel’s review and validation of each of these categories, including commentary on some of the concerns that the Panel raised and modifications made or actions taken by ENSR to address these concerns in a final draft.

4.i. Data evaluation and hazard identification

One of the first aspects of the HRA that the Health Panel addressed was the quality of the technical and health data.

The technical data used in the HRA were generated from a sampling programme carried out by ENSR in the two Barangays between April 29 and July 4, 2003. Samples were collected from the surface soil, subsurface soil and groundwater. The premise of the sampling was that some of the ‘constituents of potential concern’ that had been found onsite may have migrated offsite. These were the ‘hazards’ identified. The sampling process was designed to detect these constituents, which included hydrocarbons, lead and organochlorines (such as the family of pesticides known as ‘drins’).

The Panel had some comments and queries regarding the technical data. After careful review, and considering how vital the sampling results were to the final conclusions of the HRA, the Panel felt that the first draft did not adequately describe the sampling strategy. For example, the panel wanted greater clarification on whether and how potential pathways linking the site to the community had influenced decisions on where to locate sampling points. A Sampling Strategy Report was appended to a revised HRA which, taken together, provided the clarification the panel needed to validate this section.

In terms of health data used, the HRA contained a section of health statistics that the Panel found helpful in providing a background picture of disease patterns in Manila and separating ‘background noise’ from conditions which might be considered as risks attributable to potential exposure. However, statistics for the two Barangays and the City of Manila were not available in terms that allowed for meaningful comparison between the two populations. ENSR incorporated all relevant information and comparisons between Pandacan and Manila in the final version and the Health Panel acknowledged that the availability of good quality additional data was limited. 

The Panel also suggested that an examination of the morbidity and mortality rates of high risk groups like pregnant women and children be made so that the results could be compared with other population groups. ENSR responded that such information was not available: the health statistics obtained from the City of Manila and Pandacan are not disaggregated for pregnant women and children.  The Panel was content that ENSR had exerted its best efforts in addressing the Panel’s comments and suggestions on the issue.

4.ii. Exposure assessment

Once the sampling data has been gathered, the first step for ENSR in assessing the risk was to apply very conservative and generic ‘screening levels’ to the data. There are various international standard screening levels that can be used, and ENSR needed to decide which one best suited the conditions in Manila. After evaluation, ENSR concluded that the screening levels developed for EPA Region 9 in the USA, a region that considers exposure pathways that closely approximate the potential pathways considered in this HRA, were the most appropriate.

However for the soil-to-groundwater pathway, screening levels developed by the EPA Region 3 were used to evaluate the potential for compounds to leach from soil to underlying groundwater. These values were chosen on the basis of similar soil characteristics that are influenced by climatic conditions. Region 3 receives a much greater amount of rainfall than Region 9 and is considered to more closely approximate the climatic conditions in Manila.

The Panel had initial reservations about whether these US levels were sufficiently adaptable to the local scenario and asked for greater clarification from ENSR, suggesting that site specific screening levels be developed. ENSR explained that the EPA Region 3 soil screening levels are conservative default screening levels and that the development of on site specific screening levels would actually result in less conservative soil screening levels .The revised HRA addressed these issues in greater detail and the Panel agreed that the screening levels used were appropriate for the assessment.

After this initial phase, ENSR were able to eliminate many substances from the risk assessment as the levels found in the sampling fell below even the most conservative screening levels. The substances that remained were taken to the next stage, a more detailed risk assessment tailored specifically to local facts and conditions.

Whilst the screening levels are internationally accepted standards, the HRA carried out by ENSR made some modifications to take into account the different local conditions in Manila such as food consumption and average body weight. ENSR obtained local data from the Philippine Department of Health, the Philippine Department of Science and Technology and the Philippine Department of Environment and Natural Resources. A summary of demographic data such as population density, age and sex distribution, as well as employment data about the City of Manila and Pandacan was presented in the report.

For this second phase, ENSR also needed to establish an accurate picture of the physical characteristics of the local community and the potential exposure ‘pathways’ (primarily soil and groundwater). The process involved interviews with local residents and community representatives and the information obtained was used to develop the possible ‘exposure scenarios’ relevant to the community.

These scenarios describe the potential pathways for residues, the habits and profile of groups in the community and the activities of these groups that might expose them to these pathways. ENSR provided four groups: a typical resident of Barangay 833 or 834 (adult and child); an owner of chickens (adult and child); an adult laundry washer; and a resident urban farmer (adult and child).

The Panel suggested that an additional receptor category be added to ENSR’s list: a ‘construction worker resident’ who may be exposed both in his domestic and professional life. Whilst inhaling dust was unlikely to be an exposure pathway for the other receptor categories, as the vast majority of Barangays 833 and 834 are paved, the Panel felt dust could be generated by a construction worker’s excavation activities. The revised report contained this new category of receptor and considered the possibility of dust as a pathway for the new ‘construction worker resident’ scenario.

Because water is supplied to Barangays 833 and 834 by Maynilad Water Services, local groundwater is not used for domestic purposes and therefore a major potential pathway for residues was eliminated. However, in response to the first draft of the HRA, the Panel commented that the historical information on the two Barangays neighbouring Pilipinas Shell’s depot was insufficient.

This concern was resolved in the final report by the incorporation of statistics on historical sources of household water supply in Barangays 833 & 834, historical land use and historical improvements made to roads and pathways in the area.  This added more context to the ‘snapshot’ analysis of the local potential receptors and pathways.

The panel also recommended that an additional ‘pathway’ be added to the exposure assessment: floodwater. Flooding occurs in the two Barangays when typhoons, heavy rains, or high tides of the Pasig River occur. Research in the community found that flooding occurs approximately 50 times a year. ENSR included flooding as a pathway in its revised report although, because of a lack of data on possible concentrations of residues in flood water, this pathway was only considered in a semi-quantitative manner.

4.iii. Toxicity Assessment

The final variable to be added to the model is a toxicity assessment of the substances that the first conservative assessment identified as ‘of concern’.

ENSR analysed the approaches used by the environmental agencies in the Netherlands, UK and the US as well as those from the World Health Organisation (WHO) and the Total Petroleum Hydrocarbon Criteria Working Group (TPHWG), to establish ‘best practice’ for use in the HRA.

The Health Panel asked for clarification on how the potential toxicity of lead was evaluated. This was provided in the revised report.

4.iv. Risk Characterisation

The risk characterisation brings together the data gathered, the exposure assessment and the toxicity assessment to draw conclusions about the extent and nature of any health risk posed by the presence of residues.

For all exposure scenarios in the HRA, including the ‘construction worker resident’ scenario and the floodwater pathway, the estimates of risk came in well below the precautionary target rates. Therefore, the HRA concluded that concentrations of residues in the soil and groundwater do not pose any adverse effect on health.

Some hydrocarbons were found at a shallow depth at one test site. An assessment was carried out which determined that it was old or ‘weathered’ diesel. An analysis was then undertaken to determine whether it presented any health risk. In considering all current exposure pathways, including inhalation of vapours, it was concluded that it had no adverse effect to the health of the community.

Although there was no way of knowing how the diesel came to be at this one test location, the Panel wanted further analysis of whether it could have posed a health risk when it was ‘fresh’. ENSR thus needed to find out how old the diesel was, whether the toxicity of the diesel would have been different when it was ‘fresh’ and whether any different pathways at that time could have caused greater exposure to the local community.

Based on fingerprinting analysis, the diesel was found to be between 12 and 18 years old. A toxicity assessment showed that the risk posed by fresh diesel in the past were low relative to those based on current conditions. And discussions with residents showed that the area had been concreted and residential for some years before the diesel could have entered the ground. The diesel could therefore be discounted as both a current and past health risk.

5. Conclusions of the Health Panel

According to the Panel, an area of deficiency in the carrying out of the Health Risk Assessment was a lack of reliable health data for the community. The Panel acknowledged this limitation for sound comparison of population health status between Pandacan and Manila.

In terms of technical and environmental data, the Panel noted that it would have been preferable to have been involved from the very beginning of the project when the sampling strategy (Environmental Site Assessment) was designed. The sampling is so fundamental to the exposure assessment component of the Health Risk Assessment, that the Panel would have wanted to have had an input into its design and strategy. This would have minimised some of the Panel’s queries and requests for clarification. That said, ENSR’s responses to queries and its follow-up research and redrafting did adequately address the Panel’s concerns.

The HRA covered only those risks associated with residues in soil or groundwater. In order to get a more complete picture of any risks posed by the oil depot on the health of Pandacan residents, the Panel suggested to Pilipinas Shell that air quality monitoring be carried out. Pilipinas Shell accepted this recommendation and, at the time of writing, preliminary testing and analysis have been carried out and the panel is once again beginning its process of review.

In conclusion, the findings validated by the Health Panel indicate levels that do not adversely affect the health of the community.