OIL BEHAVIOUR AND THE RESPONSE TO A SUNKEN OIL SPILL OF SLURRY IN QUINTERO BAY, CHILE

Slurry oils, also known as catalytically cracked clarified oils (CCCO), are one of the few groups of oils that have a density greater than full salinity seawater. Consequently spills of such oils at sea are rare, but when they do occur, the response can be difficult, with low final recovery rates. Oil spill response planning and recovery techniques have been refined over many years and the research and development of most methods has been underpinned by the principle that oil floats on water. This recent case study provides an example of the behavior of a slurry spill and the subsequent response and recovery operation. Importantly, the response included systematic SCUBA SCAT (Shoreline Cleanup Assessment Technique) and ROV (Remotely Operated Underwater Vehicle) surveys to accurately document the bottom oil distribution.

A leak of a slurry occurred from a submarine pipeline in Quintero Bay, Chile, on 14^th May 2016. The oil sank to the adjacent sea bed and covered an area of approximately 10,000 m² in water depths of 18–22 m. The spilled oil had an API gravity of 4 (Table 1) and immediately sank. The majority of the oil was deposited directly adjacent to the broken pipe and the affected bottom area with the highest oil concentrations was divided into five operational work zones (Figure 1).

The incident was directed and managed by ENAP who operate the loading pipeline. The initial response involved dry-suit divers with surface air to recover the oil with vacuum hoses. These vacuums were then supplemented by four diver-operated vacuum dredges. Fortunately, weak bottom currents did not redistribute the oil; nevertheless “sea bed booms” were deployed around the perimeter to prevent migration. These booms, indicated as “barreras” in Figure 1, were initially a screen mesh that was later supplemented by sturdier metal frame screens deployed around the zone of the sunken oil in a zigzag configuration (Figure 2).

By 19 May, the initial removal phase had involved:

• approximately 20 divers using five hand-held 2” hose diameter pump systems,

• a total recovery rate on the order of 100–110 m³/day, of which an estimated 10% was oil,

• procurement of additional divers and dredges, and

• construction and deployment of sea bed barriers to contain and prevent the possible movement of the sunken oil.

Due to the water depths (16–20m) diver bottom time was limited to approximately one hour per day. Daily perimeter inspections determined whether the oil had spread farther and benthic sediment samples were collected outside of the affected area to determine background tph levels.

The initial “Emergency” Phase transitioned into a Planned Phase on 19^th May. The plan for the Planned Phase included a SCUBA survey to determine the exact location and extent of the sunken oil. The increased level of effort in this planned phase involved as many as 70 divers, operating at different times during the day due to the limited bottom time at the working depths, and the deployment of over 180m of sea-bed barriers.This phase continued until the sunken bulk oil was removed to the recommended end points and the Completion Phase involved a systematic ROV inspection to ensure that removal to the end points had been achieved.

Over the days following the release, large “cilíndros”, up to 20 cm in diameter and several meters in length, were broken from the mat by the action of orbital wave motion near the sea bed. These “rollers” were progressively further broken down into smaller individual particles.

For the purposes of mapping and documentation on this incident the following sunken oil character definitions were developed:

• mat/carpetas: 90–100% continuous distribution (Figure 3),

• cylinder/cilíndros: linear, rounded, baseball bat-type shapes detached from the mats (Fig. 4),

• globules/globulos: cohesive, rounded slurry >5cm in size (diameter or length) (Fig. 5),

• balls/pelotas de golf: cohesive, rounded slurry 1–5cm in size (diameter or length) (Fig. 5),

• particles/partículas: <1 cm in size.

The following points summarize the information generated by diver observations and by a review of near-bottom video tapes:

• AREA: Initial estimates of the extent of the oiled sea bed prior to 19^th May indicated a total oiled area on the order of 7,100m²; with additional information, this area was increased on the 19^th to 8,300m² as diver observations covered the full perimeter area. Of this total oiled area, an estimated 2,393 m² was characterized by mats or “carpetas”.

• OIL MOVEMENT: Diver observations indicated that the large mats adjacent to the pipe break point had moved only on the order of 25m over the initial 3–4 day period. This indicated that once the oil had been ejected from the pipe it moved little within and beyond the initial area of deposition over the next 5-day period (through 20 May).

• SEDIMENT INCORPORATION: The motion of the oil did not result in the incorporation or mixing of oil and sediments, although shell hash and shell fragments adhered to the surface of the oil.

• OIL CHARACTER: The orbital water motion at the sea bed by wave action was observed to lift the oil off the sea bed. The large mats (carpetas) and rollers (cilindros) moved very little vertically; the globules (globulos) and balls (pelotas de golf) were entrained up to 50cm.

By 19^th May, divers had observed:

• an estimated 2,400 m² area characterized by “heavy” sunken oil, with a total perimeter of ~175 m perimeter; this area typically had a 50%–100% oil distribution and the oil was up to 10 cm thick in places.

• this zone was characterized by either:

  • ○ a 90–100% continuous, malleable mat/carpetas typically 10cm thick or greater,

  • ○ a 20–40% distribution of parallel “rollers”, or cylinder/cilíndros, up to 10 cm in diameter and often attached and extending linearly for several meters from or in the area immediately adjacent to the continuous mat, and spaced at 50–100 cm, or

  • ○ individual 5–15 cm diameter, rounded globules.

• a larger total oiled area of 8,300 m² with a 320 m perimeter, with most of the outer areas characterized by a lower oil distribution (estimated 1–25%), and characterized by 5–15 cm diameter, cohesive, rounded globules.

A preliminary estimate on 20^th May suggested that as much as 90% of the of slurry remained in the 20% of the affected area immediately adjacent to the pipe break, with the remaining 10% distributed in the much larger (80% of the affected area) to the west and the southeast.

Penetration into the sea bed sediments did not occur as the slurry was very cohesive. From a direct sample of the slurry, the oil had a Pour Point of 9°, and the water temperature was on the order of 11°. As the oil did not physically disperse or breakdown in the low-energy sea bed environment, a semi-quantitative visual recovery target was proposed. The recommended end point was based, in part, on (a) operational practicality and feasibility, and (b) the low environmental risk associated with the oil character. The recovery target was:

• ○ no oil particles greater than 10 cm in diameter, OR

• ○ no more than one 5-cm diameter particles per meter square (< 0.5% distribution)

Operations divers had defined the perimeter of the oiled area by 20^thMay. To accurately document the distribution of the oil and focus the removal operations, a systematic SCUBA survey by oil spill observation professionals was conducted 30 May-2 June along transects positioned using AquaMap^™ transducers (Figure 7). This sea bed survey followed standard SCAT procedures, which are typically applied to oiled shoreline areas (IPIECA/OGP, 2014), and covered over 57,000m², of which 32,000m² had no observed oil.

The survey documented that oil had not spread beyond the initial affected area further into adjacent sea bed areas (Figure 8).

The key results from this survey were:

• complete mapping of the oiled area and the adjacent un-oiled areas, by professional oil observers, based on the oiling categories developed for this spill situation, using a precise positioning system

• documentation that, by 2^nd June, the most heavily oiled areas had been reduced to 382m² (25–50% distribution of carpetas – 341m²; 25–50% distributions of cilíndros – 41m²) (Figure 9).

Near the completion of the removal operations, a systematic ROV video survey with 5-m track line spacing was conducted by Marine Duty on 9^th and 10^th June. The maximum area of the mats was estimated on 20^th May to be no larger than 2,500 m². This area had been reduced by the diver removal operations to approximately 340m² by 2^nd June and to less than 50m² by 10^th June.

Marine Duty Limited was contracted to conduct an ROV video survey on 9^th - 10^th June to systematically inspect the oiling conditions after the recovery operations were considered to be completed. The survey was conducted:

• with the ROV streaming real-time video from a distance approximately 1-m above the sea bed,

• with 30 North-South transects on a line spacing of 5m; each transect was 140 m in length to cover all of the five Work Areas (Figures 1 and 9), and

• at a speed of approximately 5m/minute, that allowed real-time inspection of the imagery.

The key results from this ROV video survey were:

• a total track line of more than 4,000m that provided a continuous coverage of an area more than 20,000m²

• residual oil was identified and mapped at a total of 27 individual locations (Figure 10)

• the identification of two limited areas (one approximately 20m², the other 25m²) with small deposits of Carpetas or Cilindros (observations #12, 15, 18; and observations #16, 19 and 20) (# 16 is shown in Figure 11 left); both of these two areas had been mapped previously by the dive survey as < 1%, Carpetas or Cilindros (Figure 9)

• in the area of lines 1 through 15, all other locations were Pelotas (Figure 11 right) with one exception (observation #4: one Globulo/Cilindro)

• In the area of lines 16 through 30 that covered the eastern half of the area surveyed there was No Observed Oil, with the exception of two (2) individual locations (one Cilindro and one Pelota).

Based on this ROV survey, ENAP scheduled recovery of oil in the two Carpetas and Cilindros areas for 11^th June. This action was completed on 12^th June.

The removal operation successfully achieved the recovery of the spilled oil to the level of ALARP (As Low As Reasonably Practicable). No further treatment was recommended for the few remaining, small Partículas.

More than >95% of the oil cover documented on the sea bed was recovered over the 4-week operational period. The successful recovery of the spilled slurry was possible because of a number of oil character and environmental factors:

• the oil sank immediately and did not spread significantly, less than tens of meters for the thick mats, so that the operations activities were within a total area on the order of 10,000 m²,

• at its greatest extent during the Phase 1 bulk oil removal operations, the majority of the mat oil category covered an area no larger than 2,500 m²,

• the sea bed was within normal, surface-supplied air, diving limits; although bottom time was restricted to on the order of 1 hour,

• calm sea state and wind conditions prevailed throughout most of the Phase 1 and 2 operational periods,

• the oil could be easily recovered by standard, locally available, vacuum systems,

• recovery resources were sufficient to ensure an efficient operation, and

• systematic ROV video and diver visual surveys provided accurate data on the location and character of the oil.