Aerial Application of Herding Agents can Enhance In-Situ Burning in Partial Ice Cover

In-situ burning (ISB) offers an effective oil spill response tool in a variety of ice concentrations. The key to effective ISB is thick oil slicks. In low ice concentrations oil on water can rapidly spread to become too thin to ignite. The focus of herder research for Arctic oil spill response has therefore been on their application in drift ice conditions (1 to 6 tenths ice cover) in which slicks can spread fairly rapidly. Another potential advantage of using herders in drift ice conditions is the possibility that the entire operation could be carried out using a rapidly deployable platform such as helicopters, or possibly even remote control aircraft, to spray herders on the water around slicks and then ignite the thickened oil with aerially-deployed igniters. This type of totally aerial response could be much faster, more effective, safer and less complicated than conventional icebreaker-based countermeasures in Arctic waters. In past studies, herders have been shown to be capable of increasing freefloating slicks from unignitable fractions of a millimetre thick to ingnitable thicknesses of several millimetres. Once large herded slicks are ignited, the air drawn in by the ensuing fire can increase further thicken the burning oil and produce high removal efficiencies.

Two herding agents, ThickSlick 6535 and OP-40 have been listed on the U.S. EPA National Contingency Plan (NCP) Product Schedule for consideration for use on spills in U.S. waters. This paper describes field research to advance the operational utility of aerially-applied herders and igniters for ISB in open water (<10% ice cover) and drift ice conditions.

The primary objective of the field research was to validate the use of herders in combination with ISB, when both are applied by helicopter. The aim was to develop a rapid response aerial system that enhances responders’ ability to use offshore ISB in drift ice conditions. Specifically, the research involved five releases of 75 or 150 litres each (approximately 20 or 40 gallons) of Alaska North Slope (ANS) crude oil in a large, shallow test basin constructed on land.

The University of Alaska Fairbanks (UAF) had primary responsibility for: providing technical support, acquiring all permits required for the experiment including site approval, test basin construction, conduct of the tests, and disposal of any test materials. The test basin was sited and constructed at the Poker Flat Research Range (PFRR), an extensive land area managed by the University of Alaska Fairbanks (UAF) and situated approximately 50 km (30 miles) northeast of Fairbanks, Alaska.

The test basin was a square pool 90 metres (300 feet) on a side, contained by a 1 m (3 ft.) high lined berm (Figure 1). The interior of the berm was lined with armoured rock, and a 10 cm (4 in) wide strip of metal flashing embedded into the rocks approximately 15 cm (6 in) above the basin bottom. Both the rock and the flashing were installed to prevent scorching and contamination of the liner itself. The basin was constructed in the fall of 2014.

The experimental plan called for 10% ice coverage through the central part of the basin, leaving a 10-metre perimeter clear to allow easier clean-up of any oil and residual herder remaining after each test. As such, “ice floe” forms were placed within the basin in early winter 2015 and filled with water from a truck. In the days prior to the start of the test, the pool was filled with approximately 15 cm (6 in) of water by pumping from a nearby pond. Initial laboratory tests in 1-m² pans documented that the herders were equally effective in fresh and salt water, allowing the full-scale tests to be performed in fresh water without loss of technical integrity.

Record warm conditions over the subsequent two weeks caused significant melting of these ice floes prior to the test program. It rapidly became evident that the ice floes would deteriorate to such an extent as to be inadequate by the start of the tests, so alternative “faux” ice floes were created. Inasmuch as the main purpose of the ice floes was as a mechanical barrier to oil spreading, metal rings of various sizes were constructed to act as ice floe surrogates. Galvanized metal flashing in 2.4 m (8 foot) lengths were formed into rings using one, two, or three lengths to form circles with approximately 1 m, 2 m, and 3 m (3-, 6-, and 10-foot) diameters. Large pieces of white sorbent were placed inside the sheet metal rings to provide contrast for the aerial photography. The rings were placed providing “ice” coverage of approximately 6% in the central portion of the basin (Figure 2).

The test plan called for an instantaneous release of fresh ANS crude oil in the central portion of the basin. To achieve this, a spill release frame was constructed using 7.6 cm (3-inch) angle aluminum to form a 2.4 m (8-foot) square (3). The square was elevated by placing it on concrete paver blocks such that it had approximately 2.5 cm (1 inch) of freeboard. The selected volume of oil for each test, nominally either 75 or 150 litres, was then carefully poured into the ring (Figure 3). Stainless steel aircraft cable was attached to the frame and run to a position outside the basin at the upwind direction. When each test was initiated, workers pulled the cables to move the frame off its supports, dropping it below the water surface and causing the oil to be released.

To estimate the amounts of crude oil released for each test, the fresh ANS crude oil was pumped into numbered plastic buckets that were then weighed using a digital hanging scale. These buckets were subsequently carried out to the release frame and carefully poured into the frame using a spill plate to ensure that it did not splash or contact and stick to the basin liner. Once a burn had extinguished, the burn residue and any unburned oil was carefully collected with pre-weighed sorbent, placed in a pre-weighed plastic bag and reweighed on a digital hanging scale. The losses of oil to evaporation were not measured as they would not be significant to the mass balance in the short period of time between oil release and ignition.

Figure 4 shows the Herder Application System mounted in the helicopter used for the tests and shows the system spraying herder onto the basin. Figure 5 shows the Heli-torch, under the helicopter, being used to ignite the slick herded with ThickSlick 6535 in Test 5.

Additional details of the test data collected, data analysis and results may be found in the final report (Potter et al. 2016 - http://www.arcticresponsetechnology.org/wpcontent/uploads/2016/06/poker-flats-report-final.pdf).

This initial test involved 70 litres (18 gallons) of fresh ANS crude oil. Winds were 1 to 1.5 m/s (measured at a height of approximately 3 m above the ground). The air temperature was between 6° and 7°C and the water temperature was 5.5°C. The sky was clear and sunny. The rind of ice that formed overnight on the basin had melted and open water conditions prevailed between the real and artificial floes.

The helicopter herder application system was loaded with approximately 15 litres (4 gallons) of OP-40. Based on the results of dry runs, additional hose had been added to the system reel and all 60 metres (200 feet) of hose was deployed beneath the helicopter. To minimize rotor wash the pilot was instructed to maintain an altitude of at least 70 m (220 feet), keep forward motion at all times, and not hover over the basin.

The Heli-torch was loaded with 19 litres (5 gallons) of a mixture of 60% diesel / 40% gasoline gelled using a two part liquid-based gelling agent, Flash 21A and Flash 21B.

The oil was released from the aluminium square in the middle of the test basin when the helicopter was positioned off to the side of the basin with the herder application hose deployed and charged with herder. It was noted that the added length caused the column of herder to open the 60 psi check valve at the end of the hose, and herder dribbled from the nozzle body even when the pump was not activated. The oil was allowed to spread for a short time (10 to 20 seconds) then the pilot was instructed to apply the herder. The helicopter was used to apply two lines of herder down each side of the basin by making one upwind (inside the southeast edge of the basin) and one downwind pass (inside the northwest edge of the basin). The pilot had difficulty controlling the swing of the 200-foot hose. A total of approximately 11 litres (3 gallons) of herder was applied by the helicopter. This is in excess of the recommended dose rate (15L/km or 150 μL/m²) but such overdosing does not detract from the effectiveness of the herder.

A small amount of additional herder (1.4 litres = 0.4 gallons) was manually applied from garden sprayers to the water along the sides of the basin not sprayed by the helicopter to ensure that the test slick was entirely surrounded by herder. This was necessary because the relatively small size of the test basin severely limited the amount of time that the oil had available to drift freely before contacting an edge. Once the oil had contacted an edge, the experimental data collection part of the test was over, because it was not possible in the subsequent data analysis to distinguish between oil herded by the surfactants and burned while the slick was free-floating and oil herded by wind and burned against the metal edge of the basin.

Another artefact of the relatively small basin was that the oil was not allowed to spread to its equilibrium thickness before herder was applied, as had been done in earlier herder experiments (SL Ross and DCE 2015). Rather, for these tests, the herder was applied a short time after the slick was released and used to slow and eventually stop its spread before it reached its natural equilibrium thickness. If the body of water on which these tests were conducted was much larger, the oil could have been allowed to spread to its natural equilibrium thickness before applying herder and igniters.

As the Test 1 slick approached the edge of the basin, it was visually confirmed that the herder was acting on the slick edge: the edges were distinct and rounded with no sheen bleeding out, typical of herded slicks.

Approximately six minutes after the application of the herder had ended, the helicopter returned with the loaded and activated Heli-torch. Despite several attempts, the 60% diesel/40% gasoline fuel mixture would not ignite as it passed by the propane ignition system of the Heli-torch. The helicopter landed, and the Heli-torch was reloaded, but again the fuel mixture would not ignite. By this time the slick had drifted into the north corner of the basin and was manually ignited. After the crude oil slick had burned off, the Heli-torch was loaded with straight gelled gasoline: this ignited successfully in a brief test of the ignition system that did not involve a crude oil release.

This test was essentially a repeat of Test 1 with a different mix of gelled fuel for ignition. It involved 75 litres (20 gallons) of fresh ANS crude oil. Winds were 1.5 to 2.1 m/s. The air temperature was approximately 10°C. The sky was clear and sunny. Open water conditions prevailed between the artificial and few remaining real floes.

The helicopter herder application system was loaded with approximately 19 litres (5 gallons) of OP-40. Based on the results of Test 1, only 37 metres (120 feet) of hose was deployed beneath the helicopter. To minimize rotor wash, the pilot was instructed to maintain an altitude of at least 46 to 52 metres (150 to 170 feet), keep forward motion at all times, and not hover over the basin. The six model #0002 nozzles in the nozzle body for Test 1 were replaced with model #2504 nozzles to provide better atomization and flow of the herder.

The oil was released from the aluminium square in the middle of the test basin when the helicopter was positioned off to the side of the basin with the herder application hose deployed and charged with herder. The oil was allowed to spread for a short time (10 to 20 seconds) then the pilot was instructed to apply the herder. The helicopter applied one line of herder upwind inside the southeast edge of the basin and then circled around over the woods to make a second upwind pass inside the northwest edge. A total of approximately 4 litres of OP-40 was applied by the helicopter. The pilot had good control of the 120-foot hose in flight. The hose was then reeled in and the helicopter returned to pick up the Heli-torch. Again, a small amount of additional herder (0.5 litres = 0.14 gallons) was manually applied from garden sprayers to the water along the two remaining sides of the basin not sprayed by the helicopter to ensure that the test slick was entirely surrounded by herder.

Approximately six minutes later, the helicopter returned with the Heli-torch loaded with 19 litres (5 gallons) of recently-gelled gasoline. Although this fuel ignited as it passed the propane flame at the end of the Heli-torch, the gasoline was not gelled sufficiently to produce large blobs of burning fuel. By the time the small droplets of burning gasoline reached the water surface they had either already extinguished, or did not have enough flame left to ignite the herded crude. The helicopter landed, the Heli-torch was reloaded with another 19 litres (5 gallons) of recently-gelled gasoline, and returned. This second ignition attempt also failed to ignite the crude oil slick. By this time the slick had drifted into the snow along the southwest edge and into the west corner of the basin and was manually ignited. It was visually apparent as the slick approached the edge that the herder was acting on the slick.

For this test, the oil volume was increased to 151 litres (40 gallons) of fresh ANS crude oil and the release square was moved upwind from the centre of the basin to allow more time before the slick would reach the basin edge after release. Winds were 1.5 m/s. The air temperature was approximately 13°C. The sky was clear and sunny. Open water conditions existed between the artificial floes.

The helicopter herder application system was again loaded with 19 litres (5 gallons) of OP-40. Based on the successful herder application in Test 2, 37 m (120) feet of hose was deployed beneath the helicopter, and the pilot was instructed to maintain an altitude of at least 46 to 52 m (150 to 170 feet) and keep forward motion at all times. The nozzle body was still fitted with model #2504 nozzles.

After release, the oil was allowed to spread for a short time (10 to 20 seconds) then the pilot was instructed to apply the herder. The helicopter applied one line of herder upwind inside the southeast edge of the basin then circled around over the woods to make a second upwind pass inside the northwest edge, in the same way as the herder was applied in Test 2. A total of approximately 4 litres of OP-40 was applied by the helicopter. A small amount of additional herder (0.6 litres = 0.16 gallons) was applied from garden sprayers to the sides of the slick not sprayed by the helicopter to surround the slick. The hose was then reeled in and the helicopter returned to pick up the Heli-torch loaded with 38 litres (10 gallons) of fully-gelled 80% gasoline / 20% diesel mixture.

Approximately six minutes later, the helicopter returned and made a pass from south to north over the slick. The burning gelled fuel fell onto the slick and ignited the herded oil. A robust burn proceeded (Figure 6), in the middle of the basin, away from all the basin walls. Once the initial burn had died down the pilot made a second pass and ignited another portion of the slick. Only a weak burn was initiated. About three minutes after the second ignition, the leading edge of the ignited slick reached the edge of the basin, and the fire intensity increased as the oil was herded by wind against the edge and thickened. The remainder of the burning slick drifted to the west corner of the basin and continued to burn intensely in a small area (Figure 52) until the remaining oil extinguished. It was apparent as the slick remnants approached the edge of the basin that the herder was acting on the slick, even after burning.

The successful application of OP-40 herder followed by ignition of the slick with the Heli-torch in Test 3 resulted in the removal of approximately 70 to 80% of the free-floating ANS crude oil slick. Most of the remaining oil was consumed by burning against the metal edge of the basin: in total, 96% by weight of the Test 3 oil released was burned.

For this test, the oil volume was 155 litres (41 gallons) of fresh ANS crude oil and the release square was placed close to the centre of the basin due to rapidly changing wind direction. Winds were 2.0 m/s just prior to the oil release, but quickly dropped to 1 m/s just after the oil was released. The air temperature was approximately 12°C. The sky was clear and sunny. No ice had formed on the water in the basin the previous night and open water conditions prevailed between the artificial floes.

The helicopter herder application system had been emptied of OP-40, flushed with fresh water and refilled with 5 gallons of ThickSlick 6535. Based on the results of Test 2, 37 metres (120 feet) of hose was deployed beneath the helicopter, and the pilot was instructed to maintain an altitude of at least 46 m to 52 metres (150 to 170 feet) and keep forward motion at all times. The nozzle body was fitted with model #2504 nozzles.

The oil was released from the aluminium square in the middle of the test basin when the helicopter was positioned off to the side of the basin with the herder application hose deployed and charged with herder. The oil was allowed to spread for a short time (10 to 20 seconds) then the pilot was instructed to apply the herder. At this point the herder pump was started, but no herder came out of the nozzles. The helicopter hovered for several minutes until some amount of herder could be seen exiting the nozzles, as a slow, viscous dripping fluid. The helicopter applied one line of ThickSlick in an east to west pass, with a slight arc at the end of the pass to get herder coverage of the basin as much as possible from the air. The helicopter pilot did this first on the north side of the basin, then the south starting at the east side of the basin and moving westward to have the aircraft’s nose into the wind for both passes. Only at the very end of the second pass could significant amounts of herder be seen exiting the nozzles. It was learned later that this was the point at which the herder application system operator had engaged the air pressure purge system in an attempt to improve flow. Visual observations of the herder from the helicopter on the water showed it to be a viscous gel that did not spread over the water. Overall, very little viable TS6535 was applied by the helicopter due to the clogging of the system with gelled herder. The hose was then reeled in and the helicopter returned to pick up the Heli-torch loaded with 10 gallons of fully-gelled 80% gasoline / 20% diesel mixture. Additional herder (1 litre = 0.25 gallons) was applied from garden sprayers from the basin edge to attempt to surround the slick.

Shortly after the herder application was complete, the oil slick began to contact the base of the entrance ramp to the basin, on the northeast side. It was apparent as the slick approached the edge of the basin that the herder applied manually near the ramp was acting on the slick; however, it was not possible to confirm if sufficient herder had been applied to all sides of the slick.

Approximately six minutes later, the helicopter returned with the Heli-torch loaded with a fully-gelled 80% gasoline / 20% diesel mixture and made a pass from north to south over the slick. The burning gelled fuel fell onto the slick against the ramp and northeast side of the basin (Figure 55). The pilot made a second pass over the slick after circling around and ignited another area. A strong burn proceeded against the vehicle-access ramp and sidewall of the basin.

After Test 4 was complete, a set of bench-scale experiments with the herding agents were undertaken to determine the possible causes of incomplete herder application during Test 4. Based on these, it was discovered that when a small amount of water was added to OP-40, it almost instantly became a viscous gel. The same, but less dramatic, behaviour was noted when a small amount of water was added to ThickSlick 6535. Additional bench-scale tests indicated that diesel fuel mixed with the herders did not cause gelling: later tests indicated that iso-propyl alcohol mixed with the herders did not cause gelling either. Such gelling has not been observed when herders are applied to a water surface: the gelling reaction only occurs when a small amount of water is added to pure gelling agent.

This test was essentially a repeat of Test 4. The oil volume was again 155 litres (41 gallons) of fresh ANS crude oil and the release square was placed toward the upwind side of the basin to maximize the time the slick would drift before contacting the edge. Winds averaged 1.6 m/s, gusting to 3.3 m/s. The air temperature was approximately 15°C. The sky was clear and sunny. No ice had formed on the water in the basin the previous night and open water conditions prevailed between the artificial floes.

The helicopter herder application system had been emptied of the gelled herder the previous day (the cause of the inconsistent herder application during Test 4), rinsed with isopropyl alcohol, dried with compressed air, and refilled with 5 gallons of fresh ThickSlick 6535. Based on the results of Test 2, 36.6 m (120 feet) of hose was deployed beneath the helicopter, and the pilot was instructed to maintain an altitude of at least 46 m to 52 metres (150 to 170 feet) and keep forward motion at all times. The nozzle body was fitted with model #2504 nozzles. Good atomization was achieved.

The oil was released from the aluminium square when the helicopter was positioned off to the side of the basin with the herder application hose deployed and charged with herder. The oil was allowed to spread for a short time (10 to 20 seconds) then the pilot was instructed to apply the herder. Given the periodic wind gust conditions and wanting to surround the entire slick, the helicopter pilot lifted to 52 metres (170 ft.) and flew in a west to east pattern, with a slight arc at the start and end of the pass. The helicopter pilot flew first on the south side of the basin, and then the north starting both runs at the west side of the basin and moving eastward to have the helicopter nose into the wind. The nozzle was observed to be spraying herder at all times during these passes: approximately 4 litres (1 gallon) was applied by the helicopter. The hose was then reeled in and the helicopter returned to pick up the Heli-torch. A very small amount of additional herder (0.4 litres = 0.1 gallons - less than previous tests due to the better coverage by the helicopter) was applied from garden sprayers from the basin edge to ensure encirclement of the slick.

It was apparent as the slick approached the edge of the basin that the herder was acting on the slick.

Approximately six minutes later, the helicopter returned with the Heli-torch loaded with 10 gallons of fully-gelled 80% gasoline / 20% diesel mixture and made a pass from south to north over the slick. The burning gelled fuel fell onto the slick and ignited it. A good burn proceeded away from all sidewalls. After the first burn subsided, the pilot made a second pass over the slick after circling around and ignited another area. A second robust burn was initiated (Figure 7). Initially, burning occurred away from all sidewalls, but eventually after the second ignition, the burning slick drifted to the west corner under the influence of the wind, where some burning continued as the slick collected against the sidewall. In total, during Test 5 approximately 75 to 85% of the free-floating slick herded with TS 6535 was consumed by fire: the remainder (86% of the oil was burned in total) was consumed while the slick was herded against the side wall.

Table 1 summarizes the results of the five tests conducted.

Burn efficiencies for Tests 3 and 5 were estimated using three independent approaches (gravimetric, integrated time area, and maximum burn area). Full details are given in the final report (Potter et al. 2016 - http://www.arcticresponsetechnology.org/wp-content/uploads/2016/06/poker-flats-report-final.pdf).

Gravimetric Method. Burn efficiency via the gravimetric method was calculated after the completion of each of the five burns, and is defined as the ratio of the mass of oil burned to the initial oil mass. The following equation was used to calculate the overall burn efficiency for each experiment using the equation below.

The residue was assumed to be water free. As described previously, the mass of oil added to the basin was measured prior to each test, and the mass of residue and unburned oil was quantified after each test. Using this method, 6.0% of the original mass following the Test 3 burn was recovered, and 14% of the original mass following the Test 5 burn was recovered. Thus, the burn efficiency obtained via the gravimetric method was 94% and 86% for Tests 3 and 5, respectively. As stated previously, the wind itself effectively acted as a herding force. As a result, the gravimetric burn efficiency estimate quantifies the total oil removal efficiency by burning of not just herding agents, but also the herding action of the wind. To overcome these inconsistencies, aerial imagery approaches were employed to estimate burn efficiencies of the free floating slicks.

Integrated Time-Area (ITA) approach. In this approach, the flame area estimated from aerial imagery is integrated over the duration of the burn and multiplied by an assumed burn rate. In this instance, the burn rate of fresh ANS crude was assumed to be 1.75 mm/min, based upon a 1 to 2 mm slick thickness. While a value of 3.5 mm/min is normally assumed to be the burn rate for open water crude oil slicks 10 mm thick and greater, a lower burn rate was assumed for these tests due to the relatively thin slicks. Previous experimental data suggest that in thinner slicks, a larger proportion of heat is lost to the water underlying the slick, thus diminishing the burn rate (Buist et al. 2014).

Maximum Burn Area (MBA) approach. In this approach, that has also been used in earlier in situ burn research, the maximum burn area is determined, then multiplied by the assumed burn rate and the duration of time over which more than 50% of the maximum burn area is aflame. This is mathematically described as:

Where E₅₀ represents the time at which the burn area diminishes to half its maximum area (extinction half-time), and I₅₀ represents the time at which the spreading burn reaches half its maximum area (ignition half-time).

It is noted that the ITA and MBA approaches represent estimates only, due to uncertainties associated not only with interpretation of the aerial imagery, but also associated with the assumed burn rate of 1.75 mm/min.

During late April of 2015, in a custom-built test basin located 50 km northeast of Fairbanks AK, five tests were conducted to determine if a helicopter could be used to first apply herding agents to Alaska North Slope crude oil slicks in simulated drift ice conditions and then ignite the herded oil slicks using a Heli-torch. After some attempts failed to accomplish ignition within the time limits of the test conditions (two failed due to problems with the Heli-torch and one due to a problem with herder gelling after contact with wash water), two successful in-situ burns were accomplished using only aerial herder application and ignition. One burn was done with OP-40 herder and one with ThickSlick 6535 herder. The burning of the free-floating slicks resulted in the removal of approximately 70 to 85% of the oil on the water surface.

A number of conclusions were drawn from the testing programme including:

1. The application of two herders and subsequent ignition of free-floating oil slicks from a helicopter was successfully demonstrated.

2. The work was completed without any health, environmental, or safety incidents.

3. Both OP-40 and ThickSlick 6535 were effective in controlling the oil thickness.

4. A UAV herder application system and flare ignition system were successfully demonstrated, but additional work is needed to refine these.

5. A combined herder / igniter concept would be useful to allow for a combined herd and ignite operation in a single flight.

6. Water should not be used to rinse the herder application system as it risks gelling the herder. Isopropyl alcohol has been demonstrated to work well, and diesel has been tested at a bench-scale.

7. Further field trials of the concept of aerial application of herders and ignition in real drift ice conditions offshore are necessary in order to allow extra time for herders to act on the slick. This will allow better estimates of efficiencies and weather windows.

This paper was prepared under contract for the International Association of Oil & Gas Producers (IOGP) by S.L. Ross Environmental Research Ltd as prime contractor. Publication of this paper does not necessarily imply that the contents reflect the views and policies of IOGP, nor are there any implied IOGP endorsements of studies performed and results presented by SL Ross or by the subcontracting partners DCE and CRREL.