Alternate Retrieval Technology Demonstrations Program - Test Report

A prototype vehicle, control system, and waste and water scavenging system were designed and fabricated with essentially the full capabilities of the vehicle system proposed by ARD Environmental. A test tank mockup, including riser and decontamination chamber were designed and fabricated, and approximately 830 cubic feet of six varieties of waste simulants poured. The tests were performed by ARD Environmental personnel at its site in Laurel, Maryland, from 4/22/97 through 5/2/97.

The capabilities tested were deployment and retrieval, extended mobility and productivity, the ability to operate the system using video viewing only, retrieval after simulated failure, and retrieval and decontamination. Testing commenced with deployment of the vehicle into the tank. Deployment was accomplished using a crane and auxiliary winch to position the vehicle and lower it through the decontamination chamber, into the 36" diameter x 6' high riser, and touch down on the waste field in the tank.

The initial mobility tests were conducted immediately after deployment, prior to sluicing, as the waste field exhibited the greatest amount of variation at this time. This test demonstrated the ability of the vehicle to maneuver over the simulated waste field, and the ability of the operator to work with only video viewing available. In addition, the ability of the vehicle to right itself after being turned on its side was demonstrated. The production rate was evaluated daily through the testing period by measuring the surface and estimating the amount of material removed. The test demonstrated the ability of the vehicle to reduce the waste surface using 400 psi (nominal) water jets, scavenge water and material from the work area, and move to any location, even in the relatively confined space of the 20' diameter test tank. In addition, the ability to sluice to a remote scavenging module was demonstrated.

The failure mode test demonstrated the ability to retrieve a stuck vehicle by pulling on the tether, even if the vehicle wheels were locked or the vehicle was on its side. Line pull required to retrieve the vehicle was measured, and side load on the riser calculated from the line pull and line angles. Finally, the decontamination test demonstrated the ability to effectively clean the umbilical and vehicle.

The issues addressed and resolved during the testing were: Feasibility of deploying a vehicle-based system, mobility, production rate and limitation of water in the tank during sluicing, mining strategy, operator efficiency, vehicle recovery, and decontamination. Water usage and waste removal rates were used to estimate the time and water usage requirements for cleaning a Hanford SST.

1 INTRODUCTION

The purpose of these tests was as follows: To evaluate the capabilities of a prototype vehicle-based waste removal system using a mockup tank and test apparatus that simulated in-tank conditions; to identify and resolve design, safety and operational issues; and to identify other issues that would have to be resolved in the future.

The tests were performed by ARD Environmental personnel at its site in Laurel, Maryland, from 4/22/97 through 5/2/97. The capabilities tested were deployment and retrieval, extended mobility and productivity, the ability to operate the system using video viewing only, retrieval after simulated failure, and retrieval and decontamination.

2 TESTS PERFORMED

2.1 DESCRIPTION OF TESTS

2.1.1 System Deployment (Pre-Insertion)

Calibration of pressure, flow and level sensors was accomplished from 4/14 through 4/18. The working water pumping system, consisting of the CD100 transfer pumps, HL80 feedwater pump, and the Paco pressure pumps were also tested and flushed during that week, as were the free water scavenging module, vehicle scavenging module, and sluicer. Pre-testing commenced on Monday, 4/19 with checkout of the vehicle and control system, the data logging system, and the camera systems. The test plan and procedures were reviewed, and duties assigned to test personnel.

2.1.2 Deployment

Testing commenced on 4/22 with deployment of the vehicle into the tank. Deployment was accomplished using a crane and auxiliary winch to position the vehicle and lower it through the decontamination chamber, into the 36" diameter x 6' high riser, and touch down on the waste field in the tank.

2.1.3 Extended Mobility/Production Rate

The extended mobility testing was accomplished throughout the test period, as the vehicle was maneuvered in the waste field. However, the initial mobility tests were conducted immediately after deployment, prior to sluicing, as the waste field exhibited the greatest amount of variation at this time. This test demonstrated the ability of the vehicle to maneuver over the simulated waste field, and the ability of the operator to work with only video viewing available. In addition, the ability of the vehicle to right itself after being turned on its side was demonstrated.

The production rate was evaluated daily through the testing period by measuring the surface and estimating the amount of material removed. The test demonstrated the ability of the vehicle to reduce the waste surface using 400 psi (nominal) water jets, scavenge water and material from the work area, and move to any location, even in the relatively confined space of the 20' diameter test tank. In addition, the ability to sluice to a remote scavenging module was demonstrated.

2.1.4 Failure Mode Test

2.1.5 System Retrieval and Decontamination

This test demonstrated the ability to retrieve and decontaminate the umbilical and the vehicle, using water jets only. The volume of water required for a decontamination was also determined.

2.2 TEST METHODS AND TEST EQUIPMENT

2.2.1 Test Methods

2.2.1.1 Deployment

To gain insight into the process of deploying, operating, and retrieving the system in a tank, and aid in the identification of any critical procedural or design issues, the prototype components of the system were deployed into, operated within, and retrieved from the test tank following a set of prescribed operating procedures. The components of the system tested included the vehicle, equipped with the sluicing system, manipulator, umbilical, tether cable, and its control and power units; the sluicer scavenger module with its umbilical; the free-water scavenger module with its umbilical; and a mockup of the umbilical handling reel. The sluicer and scavenger systems were inserted and removed through the decontamination chamber by the crane. The umbilicals were passed over a cable reel to simulate the umbilical handling winch.

The objective was a better understanding of the steps that should be taken during the operation of the various components of the system in a tank. The vehicle and its scavenger module were inserted, positioned, operated, and removed in accordance with the prescribed operating procedures. In addition, the entire procedure was video taped for subsequent analysis.

2.2.1.2 Extended Mobility/Production Rate

The objectives of this area of testing were to demonstrate that the candidate system can traverse and function in the physical environment characteristic of the waste in a tank, to obtain data needed to allow the estimation of the in-tank operational time to clean a tank, and to evaluate the ability of the operator to function effectively using only video viewing. In addition, the mobility test included a demonstration of the ability of the vehicle to restore itself to an upright condition after being rolled over on its side.

The vehicle, equipped with the sluicing system, manipulator and umbilical, and the control and power units were tested. The sluicer was operate using sluicing water supplied at 100 gpm and 400 psi, while the sluicer scavenger and free-water scavenger eductors were operated at 25 gpm at 400 psi. Sluicer supply water, eductor motive water, and waste discharge water flow rates and pressures were monitored and recorded. In addition, the level of free water in the tank was monitored and recorded so that the average and peak levels could be determined in correlation with specific steps in the procedure. The material removed from the tank was discharged into dumpsters, and the water decanted off and reused. The total volume of water used to remove the simulants was determined, as was the time to effect the removal. The operations were video taped for analysis.

2.2.1.3 Failure Mode Test

This test was conducted to determine whether the sluicer system can be extracted from the tank if an immobilizing failure of the vehicle or its controls occurs. Three separate tests were run, one with the vehicle in free-wheeling mode, one with the wheels locked, and one with the vehicle partially on its side. The failure mode tests were performed at a point in the cleaning process when the vehicle was at the opposite side of the tank from the riser, and deliberately mired as deeply as possible in the clay-based simulant.

The vehicle controls were shut off and the vehicle was removed using its tether cable, which was led up the riser and over pulleys to a hydraulically driven winch located outside of the tank. A pulley was located at the lower opening of the riser and was secured via a cable to a point outside of the tank. This pulley prevented side loading on the riser that might distort it, and the pulley's cable was attached to a load cell that provided a measurement of the sideload forces involved in the vehicle's extraction. The extraction demonstration was timed and video taped.

2.2.1.4 Retrieval and Decontamination

A decontamination demonstration was performed to prove the feasibility and effectiveness of the proposed decontamination approach, to identify any critical decontamination procedural issues, and to identify hardware design improvements that would improve decontamination.

A mockup of the decontamination chamber was built and incorporated a water jetting configuration similar to that described in our proposal. A fluorescent dye tracer was mixed with the waste simulants. The vehicle was driven into a deep puddle of the clay-based waste simulant until it was well covered, and then moved back on to the solid waste surface. Additional clay-based waste was plastered on to the vehicle by hand, and allowed to dry for a period of time before commencing the decontamination.

The vehicle was then lifted into the decontamination chamber mockup and decontaminated following the prescribed decontamination procedures. The time and amount of decontamination water required to decontaminate the vehicle were measured. When the vehicle was deemed to be clean, it was moved to a location where it was thoroughly examined to determine the effectiveness of the decontamination process. The decontamination process was videotaped.

2.2.2 Test Equipment

2.2.2.1 Test Materials

The test materials used were simulants mixed in accordance with the formulas and processes defined by Hanford in appendix A3 of the RFP (see section 5, References), as follows:

First pour, wet sludge: 433 gallons of water, 7000 pounds of kaolin, mixed in the bottom of the tank. Approximately 100 cubic feet.

Second pour, hard pan #1: 914 gallons of water, 4950 pounds of kaolin, 5375 pounds of plaster, prepared in a concrete mixing truck. Approximately 190 cubic feet. Note that the mixture started to set during the pour, and only 100 cubic feet were poured, the rest remaining in the truck. This was caused by the exothermic reaction of the plaster setting up, after the concrete mixer was stopped and the delivery chute positioned.

Third pour: hard pan #2: 859 gallons of water, 4300 pounds of kaolin, 7625 pounds of plaster. Approximately 190 cubic feet.

Fourth pour, salt cake #1: 513 gallons of water, 22,400 pounds of K-Mag fertilizer, approximately 190 cubic feet.

Fifth pour, salt cake #3: 803 gallons of water, 20,080 pounds of K-Mag fertilizer. Approximately 190 cubic feet.

Sixth pour, salt cake #4: 16 gallons of water, 262 pounds of plaster, 2400 pounds of rock salt. Approximately 30 cubic feet.

The total poured volume was approximately 820 cubic feet. A fluorescing dye tracer was added to the mixing water to aid in evaluating the decontamination process, and all pours were allowed to cure for the prescribed period of time. The last two pours were arranged to provide a rugged surface for the mobility tests, and to test the ability of the system to flatten the surface. In addition, tramp material was dispersed through the simulants to provide more realistic test conditions. This material included chunks of cinder block, pea gravel, plastic sheets, and steel tapes. Photographs (1), (2) and (3) show the simulants being poured. Photograph (4) shows the simulant surface after the final pour.

In the course of performing the project, ARD Environmental forwarded simulant samples for analysis by Pacific Northwest National Laboratory (PNNL)personnel. PNNL raised a question concerning the particle size of the K-Mag simulant. ARD Environmental used a coarse grained material based on the vendor's information and the use of the descriptive word "fertilizer" in the PNNL specification (and used the same material for phase 1 testing). PNNL informed ARD Environmental that although the specification called for "fertilizer"grade, the fine grain material was intended. This raised the issue of the strength of the simulant as a function of particle size. At this writing, the sample analysis has not been conveyed to ARD Environmental, however the simulants in question proved to be extremely hard, as evidenced by the removal rates.

2.2.2.2 Test Equipment

Site Layout

The site layout is shown in figure 2.2-1. The tank mockup was assembled alongside the rear of the building, and the water and waste containers arranged along the same wall. The transfer pumps and feedwater pump were mounted on top of the containers, and the pressurizing pumps were located on the ground nearby. The remainder of the equipment was located in the immediate vicinity, and all hoses and piping located between the equipment and the building wall.

The crane and forklift were aligned with each other and the center of the decontamination chamber and riser assembly. The forklift served as the mount for the winch used to raise and lower the vehicle, via cable and blocks fixed to the crane boom. The crane hook was used to raise and lower the vehicle and the vehicle water scavenging module. The free water scavenging module was positioned in the tank at approximately the 2 o'clock position in the figure.

The site layout is shown functionally in figure 2.2-2, which indicates the locations of the pressure and flow sensors. All transfer and pressurizing pumps were electrically operated; the scavenging module pumps, which were submerged, were driven hydraulically.

Tank Mockup

The test tank, shown in figure 2.2-3, is a 20' diameter by 5' deep steel structure that was partially filled with mixed simulants to a depth of 3' to provide a test volume of approximately 830 cubic feet. The tank was enclosed in a scaffolding and truss structure, decked over to provide a simulated tank top 20' above the level of the tank bottom. The deck was plywood, sealed to exclude rain, and the scaffolding was equipped with OSHA rated guard rails and stairway. A grid of holes on a 2' spacing was drilled in the deck to allow soundings to be taken and the surface topology of the waste evaluated. The layout of the grid is shown in figure 2.2-4. Grid intersections were numbered serially in serpentine fashion for convenience in sounding the waste surface. Heavy duty tarpaulins were hung from the truss to enclose the tank in a curtain which to exclude rain, and contain spray and material from the tests, and further simulate operation in a tank. The tank is shown under construction in photograph (5), and the scaffolding during erection in photograph (6).

A mockup riser, 36" in inside diameter and 6' long was installed extending from the deck downward, as shown in photograph (7). A functional decontamination chamber was mated to the upper flange on the riser. Installation of the decontamination chamber is shown in photographs (8), (9) and (10). The tank also included four vertical pipes welded to its bottom to simulate risers and other obstructions that might be found in an actual tank; these pipes were extended to the scaffolding deck with plastic pipe. The operator used video monitors to guide the equipment, and did not have direct visual contact with the interior of the tank. Cameras were located at four positions, 90° apart and 18' above the level of the tank bottom. The camera diametrically opposite the riser was equipped with a pan and tilt, while the other cameras were hard mounted. All four cameras were equipped with zoom lenses. The ability of the operator to control the vehicle under these conditions was evaluated.

Sluicer and Eductor Water Supply System

The water supply system consisted of all of the components of the test setup involved in providing sluicing water to the sluicer and motive water to the two eductors used in the tank. A block diagram of the system is provided in figure 2.2-5. Four containers (sealed dumpsters) were used to recycle water during the tests. Two were designated waste tanks; one being used while the other was emptied. The other two tanks were on line permanently as water supply tanks.

Discharge from the test tank was decanted to successive containers using transfer pumps, and the last container was the source of feed water for the sluicer and eductors. Successively finer strainers were used between the containers to filter solids suspended in the water. The total area of each strainer was about 24 square feet, and the mesh was large enough to prevent clogging. The pumps used to supply sluicer and eductor water would pass any particles that could pass through the finest mesh strainer (1/16"), and no difficulties were experienced with suspended kaolin or plaster particles. The strainer boxes are shown in photograph (11).

The tanks were filled initially, and refilled as needed, from a municipal water hydrant. A water meter and backflow check valve was included in this line as required by the county. The main feed pump drew feed water from the supply tank at a rate of about 300 gpm, the actual value was measured by flow meter Q1. A flowmeter installation is shown in photograph (12), and the water supply tanks, transfer pump and feedwater pump in photograph (13). To increase the pressure of the flow to the desired level, two main pressure pumps were plumbed in parallel.

A pressure sensor, P1, was placed in the main feed line at the distribution manifold. This manifold divided the feed water into three separate lines, one to the sluicer input, one to the sluicer eductor motive water input, and one to the free-water eductor motive input. The pressures at these three points was monitored by pressure sensors P2, P3, and P4, respectively. The combined outlet size of the sluicer nozzles and the eductor nozzle orifice sizes were chosen to distribute the feed water among the sluicer and two eductors at the rate of about 100 gpm at 400 psi for the sluicer, and 25 gpm at 400 psi (nominal) for each of the eductors. The valve assembly is shown in photograph (14).

Waste/Water Scavenging System

The waste/water scavenging system consisted of the components of the test setup involved in removing waste and water from the test tank. Waste broken up by the sluicer vehicle, sluicing water from the vehicle itself, and free water in the tank were collected and conveyed topside using ARD proprietary pumping modules. For the purpose of this test, the vehicle and two waste scavenger modules were employed. One module pumped waste and water scavenged by the vehicle sluicer head, while the other scavenged waste and water that escaped from the sluicer.

Waste water was conveyed through 3" flexible hoses that passed through the decontamination module attached to the riser. The flows were measured by flow sensors Q2 and Q3 before passing through valves that were placed in the system to permit adjustment of back pressure to simulate the 55' head specified for the waste discharge lines by Hanford. The dynamic losses in the discharge hose amounted to another 18' of equivalent head. During initial system checkout , it was determined that the back pressure due to line losses was 35 psi (79') without throttling, therefore the valves were left fully open during the tests. The discharge pressures were continually monitored by pressure sensors P6 and P7. The waste water was emptied into the waste tank where waste settled out, and the water transferred to the water supply tanks and recycled.

Decontamination Water System

The water supply and instrumentation for the decontamination portions of the test are shown in figure 2.2-6. Water was obtained for decontamination from the municipal water hydrant. The amount of water used for decontaminating the system was computed from the measured flow rate and time, and was checked roughly by measuring the level of water in the tank mockup after completion of decontamination.

The water was filtered, and flow rate was measured using flow sensor Q4. A commercial high-pressure washer-type pump was used to drive the water. It provided water at the decontamination chamber at a rate of 40 gpm (nominal) at approximately 3000, 2000, and 1000 psi, depending on how many spray rings were on at any time. The pressure was monitored by pressure sensor P8 at the decontamination module. The flow could be directed to any of the three spray rings or the manual spray wand by a set of valves at the decontamination chamber.

Decontamination Chamber and Riser Assembly

The decontamination chamber consisted of a 71.5" inside diameter by 84" high aluminum cylinder with an integral flange welded to the bottom. It was bolted to the top of the riser, which had an inside diameter of 36" and was 6' long. The assembly is shown in figure 2.2-7.

The decontamination module contained three spray rings around its inner circumference, located near the bottom of the chamber. Each ring had eight spray nozzles, equally spaced around it, spraying in toward the center axis of the cylinder. The nozzles of the top ring point downward at 45°; those on the center ring straight inward; and those on the lower ring upward at 45°. The three-ring stack was mounted on a swivel that allowed 45° of rotation, enabling the spray to be directed at any portion of the unit being decontaminated. The spray ring assembly is shown in figure 2.2-8, and photograph (15).

In addition to these fixed spray rings, the chamber included two rings of eight holes, equally spaced around its circumference to allow the use of a manually operated spray wand. The holes were plugged when not in use. Components were moved vertically within the spray field by raising and lowering the component using the handling crane. Two sets of four observation windows equally spaced around the chamber were also provided. The holes for the wand, and the windows, were located one quarter and three quarters the way up the cylinder. The interior of the chamber is shown in photograph (16).

Vehicle and Sluicing Assembly

The vehicle and sluicing assembly are shown in figure 2.2-9. The vehicle is driven hydraulically, and has a maximum available torque of about 4800 foot-pounds. The vehicle weighs approximately 2000 pounds, and is equipped with hydraulic actuators to permit positioning of the sluicer over a wide range, as shown in the figure. This permits the use of the sluicer to excavate beneath the vehicle, as well as washing down walls and other structure up to the height of the sluicer in the fully raised position. The vehicle is shown in photograph (17) during construction, and the control station in photograph (18).

The sluicer is shown in figure 2.2-10, and in photograph (19), during test. It is equipped with 14 sluicing nozzles, with a nominal diameter of 0.140" and a measured diameter of 0.128", and a 0° spray angle. These nozzles deliver 114 to 122 gpm at 400 psi and 450 psi, respectively. Nozzle performance data are shown in figure 2.2-11. The sluicer is equipped with an eductor, located in the center of the sluicing nozzles. The eductor has a suction capacity of 100 gpm, therefore it is capable of scavenging all the water delivered by the sluicing nozzles, when conditions are favorable.

2.2.2.3 Instrumentation

The instrumentation that was used for this test effort is delineated in Table 2-I, below.

Table 2-I. Test Instrumentation.

Item	Description	Sensor Accuracy	Measurement Accuracy	Vendor	Part Number
P1	Pressure Transducer, 0-500 psig	±0.5% FS (±2.5 psig)	±1.7% @ P_op (±6.6 psi)	NOSHOK	100-500-1-1-2-2
P2	PressureTransducer, 0-500 psig	±0.5% FS (±2.5 psig)	±1.7% @ P_op (±6.6 psi)	NOSHOK	100-500-1-1-2-2
P3	Pressure Transducer, 0-500 psig	±0.5% FS (±2.5 psig)	±1.7% @ P_op (±6.6 psi)	NOSHOK	100-500-1-1-2-2
P4	Pressure Transducer, 0-500 psig	±0.5% FS (±2.5 psig)	±1.7% @ P_op (±6.6 psi)	GNEISSIC	100-500-1-1-2-2
P5	Pressure Transducer, 0-500 psig	±0.5% FS (±2.5 psig)	±1.7% @ P_op (±6.6 psi)	GNEISSIC	100-500-1-1-2-2
P6	Pressure Transducer, 0-150 psig	±0.5% FS (±0.75 psig)	±3.3% @ P_op (±4.2 psi)	GNEISSIC	100-150-1-1-2-2
P7	Pressure Transducer, 0-150 psig	±0.5% FS (±0.75 psig)	±3.3% @ P_op (±4.2 psi)	GNEISSIC	100-150-1-1-2-2
P8	Pressure Transducer, 0-5000 psig	±0.5% FS (±25 psig)	±1.9% @ P_op (±56.5 psi)	GNEISSIC	100-5000-1-1-2-2
Q1	Flow Sensor, 0.7-30 fps	±0.2 fps	±11.4% @ Q_op (17 gpm) uncal, ±1 gpm cal'd	Omega	FP-6000
Q2	Flow Sensor, 0.7-30 fps	±0.2 fps	±11.4% @ Q_op (17 gpm) uncal, ±1 gpm cal'd	Omega	FP-6000
Q3	Flow Sensor, 0.7-30 fps	±0.2 fps	±11.4% @ Q_op (17 gpm) uncal, ±1 gpm cal'	Omega	FP-6000
Q4	Flow Sensor, 0.7-30 fps	±0.2 fps	±11.4% @ Q_op (17 gpm) uncal, ±2 gpm cal'd	Omega	FP-6000
L1	Level Transmitter, Float Type	1/4" resolution	±1/8" max cal'd	Gems	XT-800 Type 3 SS w/ Buna N float
L2	Level Transmitter, Float Type	1/4" resolution	±1/8" max cal'd	Gems	XT-800 Type 3 SS w/ Buna N float
L3	Level Transmitter, Float Type	1/4" resolution	±1/8" max cal'd	Gems	XT-800 Type 3 SS w/ Buna N float
L4	Level Transmitter, Float Type	1/4" resolution	±1/8" max cal'd	Gems	XT-800 Type 3 SS w/ Buna N float
Vd	Water Meter	0.1 gal/gal	0.1 gal/gal	Rockwell
R	Load Cell	0-2000 lbs		Tri-Coastal	Model 264-202
A&B	Plumb Level	1° resolution	±5°	Starrett	No. 12
Wr	Scale	0.1g resolution	±0.1g	Ohaus	Triple Beam

The pressure, flow rate, and tank water level sensors, shown in photograph (20), were automatically monitored and recorded by the data acquisition system. The sensors chosen provided 4 to 20 mA outputs, which allowed the sensors to be placed hundreds of feet away from the data acquisition system. The current outputs were converted to voltages on a signal conditioner board designed and built by ARD Environmental. The resulting sensor voltage outputs were connected to a National Instruments AT-MIO-64E-3 A/D PC board. The A/D board was configured in a differential input mode for all channels.

Labview software was used to control the A/D board and scale and record the sampled data. A single software Virtual Instrument (vi) was written to complete the data acquisition, scaling, display and recording tasks. All sensor channels were sampled at a one second rate. Each channel was scaled using a gain factor and offset parameter. These scale factors were embedded in the code so they couldn't be altered during data acquisition. Each sensor output was displayed to the operator on a numeric software display in the appropriate engineering units (psig, gpm or inches). All test data were recorded in a time stamped omma separated value (csv) ASCII file format. This csv file can be read by many spread sheet programs or by Labview software.

The flow sensors and standing water level sensors were calibrated prior to testing, the former to +/- 2 gpm, and the latter to +/- 1/8". Factory calibrations were utilized for the pressure sensors, electronic levels, and load cell.

2.3 TEST RESULTS

The test series as originally planned was modified because of repairs required during the test series, and the terms and conditions of the contract. However, nearly all of the major objectives of the test series were met. Table 2.3-1 summarizes the tests, test parameters, and data derived.

Table 2.3-1: Test Description and Data Derived

Test Phase	Test Focus	Test Parameters	Key Data Derived
1. System Deployment (Pre-Insertion)	- Procedures for preparing system for tank entry.	- Pre-insertion procedures.	- Operating procedure requirements - Video record of operations.
2. System Insertion and Positioning	- Procedures for insertion & positioning of equipment in tank.	- Deployment procedures. - Impact on riser.	- Operating procedure requirements - Insight into umbilical management. - Video record of operations.
3. Extended Mobility / Production Rate Note: The testing was limited to partial removal of material because of time and equipment constraints. Approximately 11.9% of the total waste volume was removed.	- Mobility around rugged surfaces. - Extended operation. - Determination of possible production rates. - Ability to operate in the presence of debris such as steel tapes, plastic sheets, cinder block fragments.	- Starting and ending volumes of simulant. - Time to remove specified amount of simulant. - Standing water levels in tank. - Sluicer & eductor water flow rates. - Sluicer & eductor operating water pressures. - Sluicer scavenger and free water scavenger discharge rates.	- Average cleaning rate of system based on data taken. - Effectiveness of sluicer-mounted scavenging system. - Average & peak standing water in tank. - Average sluicer and eductor water usage. - Average total water usage. - Operating procedure requirements - Insight into umbilical management. - Ability to deal with cinder block fragments and sheet plastic. - Video record of operations.
4. Failure Mode Demonstrations	- Extraction of failed vehicle.	- Cable side loads on riser. - Tether cable tensions. - Relative cable angles.	- Ability to recover disabled vehicle. - Effectiveness of failure mode extraction method. - Identification of vehicle design improvements. - Impact on riser. - Video record of operations.
5. System Retrieval and Decontamination	- Procedures for retrieving equipment from tank. - Ability and effectiveness of proposed decon approach.	- Presence of dye tracer. - Volume of decon water used. - Time to complete decon. - Weight of matter remaining after decon. - Decon water flow rate. - Decon water pressure.	- Operating procedure requirements - Insight into umbilical management. - Impact on riser. - Amount of simulated waste remaining on each component after decontamination. - Volume of water used to decontaminate each component. - Effectiveness of methodology. - Identification of dose traps on each component. - Video record of operations.
Items not addresses due to modification of the test plan to meet contractual requirements	- Demonstration of complete removal of bulk waste. - Ability to clean the tank bottom to a minimum level of residual waste. - Effectiveness in dealing with large sheets of plastic, steel tapes and other miscellaneous debris.

2.3.1 System Deployment (Pre-Insertion)

The pre-insertion effort consisted of a review of the test procedures with all test personnel to confirm layout of the equipment, and the duties of the test personnel. When deployment commenced, the forklift, to which the vehicle winch was mounted, had to be repositioned to reduce side load on the crane boom. This was done, and no difficulties were experienced during the actual deployment.

2.3.2 Deployment

Deployment took place on Tuesday, 4/22. The vehicle, which weighed in at 1864 pounds, was lowered into the 36" riser using the winch mounted on the forklift. The sluicer tilt arm had to be adjusted once for clearance, and the vehicle was then lowered to the waste surface without incident. The vehicle operator performed all operations remotely using video only, which was a major parameter of the testing. The vehicle umbilical and hose were led over a large drum suspended for the end of the crane. The end of the sluicer eductor discharge hose was retained at the top of the decontamination chamber, for subsequent connection to the Vehicle Water Scavenging Module (VWSM).

The VWSM was lowered into the decontamination chamber, and then into the riser. The vehicle umbilical and hoses were also in the riser, but did not interfere with the insertion of the VWSM. The sluicer eductor discharge hose was connected to the VWSM, and it was then lowered the rest of the way to waste surface, using the hook on the crane. Note that the connection from the discharge hose to the VWSM would have to be made remotely in a deployed system.

Handling of the vehicle and the VSWM was observed by the crane operator via video camera mounted on the end of the crane boom. This image was also available to the system operator at the control console, and was videotaped. The system operator observed the deployment procedure from the control station via the video cameras installed in the tank mockup, and the main monitor video was also taped. One tank camera failed immediately prior to the deployment. This was traced to a loose wire, which was repaired immediately after deployment.

The system operator performed all manipulations of the vehicle during the deployment. This included positioning the sluicer to clear the riser, and to shift the center of gravity of the vehicle at touch down. The vehicle was driven forward after touch down to allow the VWSM to be inserted fully. After observing the arrangement in the tank, it was decided to suspend the VWSM several feet above the waste surface during operations. This was necessitated by the relatively confined space of the mockup, and may have to be done from time to time in a Hanford tank to permit removal of waste from beneath the module. However, this does not appear to pose an operational problem.

The deployment sequence is shown in photographs (21)-(25), commencing with preparation of the vehicle, the lift and insertion through the decontamination chamber, emerging from the riser, touchdown on the simulant surface, and finishing with insertion of the VWSM.

2.3.3 Extended Mobility/Production Rate

Narrative

Extended mobility and production rate testing was started on Wednesday, 4/23. The initial testing consisted of driving the vehicle around the tank to observe its characteristics. Maneuvering was not difficult, and the vehicle could easily skid steer, especially on the hard saltcake. The primary impediments to maneuverability were the small size of the tank and the umbilical, especially the high pressure water supply hoses and the discharge hose from the sluicer eductor to the VSWM. Umbilical management will need to be improved for long term in-tank operations, primarily by the use of powered handling winches with constant-tension drives to control the amount of umbilical in the tank.

During the initial mobility tests the vehicle slid sideways off of a mound of saltcake # 4, and wound up with a mockup riser stuck between the front and rear wheels on one side of the vehicle. We believe the vehicle could have freed itself had the full range of rotary motion been available between the front and rear axles. Unfortunately, a minor design error resulted in mechanical interference that limited the range of motion to ±30°, instead of the design value of ±55°. In any event, shortening the distance between the wheels would eliminate this as potential failure mode. Recovery was accomplished using a "come-along" to pull the vehicle away from the mockup riser. Mobility in the tank is shown in photographs (26), (27), and (28).

After the initial mobility tests, material retrieval commenced. The vehicle was used to mechanically dislodge material, as well as sluicing, and removing material with the VSWM and FWSM. A pump failure was experienced in the Vehicle Water Scavenging Module (VWSM) late in the afternoon of 4/23. This was due to the sealing discs in the pump jamming on the pea gravel dispersed into the simulant. The sealing discs were replaced with a harder formulation material on Thursday, 4/24, and testing continued. At the end of testing on 4/24 a vehicle motor seal was observed to be leaking. This was traced to restricted flow in the case drain lines. Larger lines were installed and the motor seal replaced the following day. Case drain flow rates were also checked and discussed with the manufacturer, who indicated that the rates were within specification.

The rest of 4/24, and all of Friday, 4/25, were devoted to maintenance of the system. The flow meters in the waste discharge lines had exhibited erratic performance, and one failed completely. The flow sensors were removed and cleaned, and the failed unit was found to be jammed by a piece of foreign material between the paddle wheel and housing. The units were tested, and then reinstalled. The hydraulic fittings on the forward end of the sluicer lift cylinders were found have been damaged by the rotating couplings on the water supply hoses to the sluicer. The cylinders were repositioned and the fittings and hoses replaced. The umbilical was also redressed and the broken tie wraps replaced with steel hose clamps. In addition, 1/2" case drain lines were run to replace the existing 1/4" lines.

A protective bracket was installed over the front of the sluicer lift cylinders and testing resumed on Monday, 4/28. The Free Water Scavenging Module (FWSM) fill rate was observed to be reduced, indicating poor scavenging. In addition, the pressure sensors on the vehicle again became erratic, and the cables were serviced and redressed. The FWSM was first raised about 4" as it appeared that waste accumulation around the base was the cause of reduced scavenging performance, however, this had no effect.

The FWSM was disassembled on 4/29, and a triangular piece of stone, approximately 0.26" across, was found lodged in the eductor motive water nozzle. This stone must have been trapped in one of the rented pumps or associated hoses, despite the extensive flushing performed prior to the testing, as the water system intake screen is 1/16" mesh. The FWSM was reassembled, reinserted in the tank, and testing continued, with the FWSM operating properly. Note that the water level in the tank reached a maximum of 10" while the FWSM was not functioning properly, primarily due to the free use of sluicing water by the operator, and the limited vehicle scavenging performed by the operator during this period.

Maintenance was performed on 4/30 to repair a hydraulic leak on the vehicle hydraulic manifold. In addition, the temporary lift cylinder rod bracket fabricated at ARD Environmental prior to the start of testing was replaced with the permanent unit, which had finally arrived from a machine shop which shall remain nameless, and never used again.

Because of the need to complete testing by Friday, 5/2, it was decided that the hard saltcake would be removed manually in an area, and the vehicle operated in the hardpan and soft sludge to determine performance in those waste simulants. Accordingly, on Thursday, 5/1, a region of saltcake was loosened with a jackhammer and removed, and the vehicle driven onto the hardpan. The area cleared was only about the size of the vehicle. When the vehicle was maneuvered onto the hard pan, it sank in fairly easily, and could not move forward because of the tank wall, nor backward because the of the umbilical stinger.

Nonetheless, the vehicle was able to sluice and remove most of the waste in the immediate area, including waste beneath it. We believe that with design changes to permit the umbilical to lead upwards, instead of straight back, and prolonged operation, a vehicle in this circumstance could gradually remove enough waste to permit motion, and eventually free itself. Testing halted late in the day when the pump in the FWSM jammed. Upon examination it was found to have seriously distorted discs, again, these were the older style softer discs.

A final mobility test was conducted to determine if the vehicle could recover its footing if it was on its side. The vehicle was rolled over onto its right side on the saltcake. The sluicer was then move to its lowest position using the lift arms, thus shifting the Center of Gravity (CG) to the left. The rear axle was then rotated to the left, further shifting the CG to the left. At this point the vehicle tipped to the left. The sluicer was then shifted to the right (up), and the vehicle rolled back onto its wheels.

Test Results

When all systems were operational, the design goals were met. The system, including the transfer and pressure pumps, could be operated by a single person at the remote control station. The water and waste scavenging systems were effective in removing waste and water, and in limiting the amount of free water in the tank. A key factor in limiting the water is developing a mining strategy that results in flow of the waste and water to the scavenging module, which must be situated at a low point.

The vehicle eductor was effective in removing pooled water and waste wherever these were encountered, however, it was generally not possible to remove water as it was injected by the sluicer. Any ventilation of the shroud would result in air being taken up by the eductor, and not water or waste. When the shroud was buried in the waste, and angled properly, it could remove sluicing water almost as fast as it was injected.

The operation of the sluicer and sluicer eductor is shown in figures 2.3-1 and 2.3-2. These figures show sluicer and sluicer eductor pressure and flow, and the cycling by the operator. Operation of the VSWM and FSWM are shown in figures 2.3-3 and 2.3-4. The automatic cycling of the pumps in the two modules indicates that the modules are functioning properly. Cycle times vary somewhat for the FWSM, and greatly for the VSWM, because the FWSM was at a low point with waste and water flowing to it most of the time.

The cycling of the sluicer and eductor water supplies for the five material removal periods is shown in figures 2.3-5 through 2.3-9,. The X-axis is test time and the dark band indicates when the water supply is was on. The sluicer water supply shows the most activity, as it was turned on and off frequently by the operator during the testing. At times the vehicle was used primarily as an excavator, and at those times the sluicer and sluicer eductor were both turned off. The FWSM eductor was on almost all of the time.

The waste surface contours are shown in figures 2.3-10 through 2.3-14. The initial surface is shown in figure 2.3-10, and the reduced surfaces are shown in the remainder of the figures. The figures accurately reflect the topography. In the first figure, the central mound is the last pour, saltcake #4. There is a gully below and to the left of the riser, and there are two terraces above the riser proceeding counterclockwise. The FWSM was located in a pit at about 8 o'clock on the figures, against the tank wall.

The productivity data are summarized in the table below, however, the text that follows contains important additional information that bears on the interpretation of the data, and should be read carefully.

Table 2.3.3-1: Productivity Data

Run # /Material	Total Time	Sluicer Water Time	Sluicer Eductor Water Time	Free Water Eductor Water Time	Waste Volume Removed	Sluicer Water Used	Time/Cubic Foot Removed	Sluicer Water Volume / Sludge Volume Removed
	Minutes	Minutes/%	Minutes	Minutes	Cubic Feet	Gallons	Minutes	Volumetric Ratio
1/SC4	101.8	33.9/33.3	66.6	80.6	52.4	4746	1.94	12.11
2/SC3	111.5	22.2/19.9	98.1	100.1	10.85	3108	10.28	38.30
3/SC3	200.1	58.0/29.0	159.9	164.3	11.9	8120	16.82	91.22
4/SC3	134.1	83.7/62.4	127.1	132.6	16.19	11,718	8.28	96.76
5/HP,SS	38.1	13.8/36.2	24.3	36.7	20.6	1932	1.85	12.54
Total	585.6	211.6/36.1	476	514.3	111.94	29,624	5.23	35.40

The data were extracted from the Labview files, and the last two columns calculated from the data. The "water" times were derived from the appropriate pressure sensor outputs, and total time from the pressurizing pump pressure sensor. The volumes of material removed were computed by the topographic software used to create the contour charts, except for run 5. This run was done in the hardpan and wet sludge deliberately exposed by manually removing the saltcake overburden. No waste surface readings were taken for run 5 since the removal was done in a localized area. The volume of material removed was estimated at 2.5' wide x 55' long x 1.5' deep based on the dimensions of the vehicle, the area exposed by removing the overburden, and the hole left after the removal. The water injected into the tank via the sluicer was derived from the average sluicer operating pressure, nozzle flow based on manufacturer's data, and sluicer operating time.

Run 1 shows a great deal of material removed for the amount of water and operating time, consistent with the nature of saltcake #4. Run 2 shows a significant increase in the time and water required per cubic foot of waste removed, due to the hardness of saltcake # 3. The vehicle was used as an excavator, and while it was able to tear up the surface of the waste, this took some time. Modifying the cleat design would improve the excavation capability.

Runs 3 and 4 exhibited a further increase in time and water used per cubic foot of material removed. This is primarily due to the greatly increased percentage of sluicer "on" time, compared to the previous run. This reflects the approach of the individual operator and the state of the equipment at the time. In this run, the operator chose to sluice extensively, and the FWSM was not fully functional due to the pebble trapped in the eductor nozzle.

It was noted that the saltcake softened during a run by excavation with the wheels would reharden overnight to its original consistency. Thus removal volume would have improved had the softened material been sluiced and removed at the end of each day. Whether or not the Hanford tank wastes exhibit this characteristic is not known.

Run 5, which was in the relatively soft hardpan and sludge shows time and water use consistent with run 1, also in softer material.

Based on observation, the maximum amount of standing water in the tank was no more than 30-50 gallons, when the FWSM was fully operational. The amount of water in the tank could easily be limited by careful operation of the sluicer, and use of the sluicer eductor to remove trapped pockets of waste and water. The maximum amount of water in the tank was estimated at 400-500 gallons, during run 4. This occurred when the FWSM was not operating properly, and the system operator kept the sluicer on for extended periods.

The average performance against hard and soft materials are shown in Table 2.3.3-2. Run 4 was eliminated from the calculation, since the FWSM was not fully functional, and excess water was deliberately used. Note that the "soft" material includes low strength saltcake as well as hardpan and sludge.

Table 2.3.3-2: Average Performance Against Hard and Soft Simulants

Simulant	Total Time	Total Volume	Total Water	Time per Cubic Foot	Water / Sludge Ratio	Percent Solids by volume
Soft: SC4, HP, SS	139.9	73	6678	1.92	12.23	8.17
Hard: SC3	311.6	22.75	11228	13.7	65.98	1.52

These averages are realistic in that they reflect real operational time exclusive of breaks or downtime, but inclusive of production inefficiencies such as vehicle maneuvering, camera adjustment, reading of gauges, in-process discussions, and the like. Assuming that the simulants are representative of the characteristics of the material in the tanks, an estimate may be made of the time required to clean a Hanford tank.

Assume the following:

Heel thickness = 6", over entire tank bottom = 2209 cubic feet.
Soft sludge = 4' over entire bottom = 17671 cubic feet.
Medium saltcake = 1.5' in a single layer on top = 6627 cubic feet.
Two 8 hour shifts per day, 5 days per week, 75% productivity = 12 hours/day, or 60 hours per work week.

The time require to remove the hard heel is about 504 hours, or 8.4 calendar weeks. The time to remove the softer materials is about 778 hours, or 13 calendar weeks. The total removal time is thus 21.4 calendar weeks or about 5 months. The total sluicing water used is about 3.3 mgal, at an average rate of 43.1 gpm. Another 3.85 mgal of eductor motive water would also be required, at a rate of 50 gpm. Although the total volume of water is large, the total average rate is less than 100 gpm, which is entirely consistent with the use of recirculated working water.

Operation of the vehicle using the video system was relatively straightforward, although only one of the tank cameras was equipped with a pan and tilt. It would definitely be desirable to have multiple tank cameras, all equipped with zoom, pan and tilt, since this would permit the operator to view the work face no matter where the vehicle was in the tank. Fatigue did not appear to be a factor, probably because of the novelty of the situation and the fact that the longest run was only 3 hours and 20 minutes.

Standing water level sensors were installed at four locations in the tank in order to monitor residual water. These sensors proved to be unresponsive to anything but long term changes in level, and were most affected by the permeability of the waste. The standing water levels over the life of the test effort, about 9 days, is shown in figure 2.3-15. Note that at the end of the test, the drain valve for L1 was opened, and the water drained from the sensor column. No additional water drained from the tank, confirming that the sludge at the bottom of the tank was essentially clogging the flow and only allowing slow permeation. The sudden drop in level at L4, which occurred over a period of 18 minutes on 5/1, is unexplained, however the most likely reason is inadvertent cracking of the drain valve.

2.3.4 Failure Mode Test

The failure mode test was conducted on 5/1. Three tests were performed, the first with the wheels in free-wheel mode, which is the most realistic condition, the second with the wheels locked, and the third with the vehicle on its side with the wheels locked. In all cases, the vehicle was driven into the hole created in the hardpan opposite the riser during the last productivity test on 5/1. The test layout and geometry are shown in figure 2.3-16. Photograph (29) shows the vehicle during a retrieval test.

The measured loads and angles, and the resulting forces are tabulated in the table below. It appears that the principal component is the vertical lift of the vehicle, thus there is little difference in loads for the three tests. Although angles and loads were not measured for the third run, the winch pressure was monitored and was in the same range as the other two tests.

Table 2.3.4-1: Retrieval Loads

Test Condition	Angle "A"	Angle "B"	Measured Load	Winch Pressure	Line Tension	Vehicle Horizontal Load	Vehicle Vertical Load	Riser Horizontal Load	Riser Vertical Load
Free Wheel	18.5	0.7	1260	1000-850	2061	1220	1015	1195	400
	19.2	0.4	1100	550	1707	1051	867	1039	362
	8	1.4	800	500	3480	877	774	792	111
Locked Wheel	17	0	1250	900	2138	1195	1036	1195	365
	21	0	1200	NM	1674	1120	892	1120	430
	20	0	530	NM	775	498	406	498	181
Tipped/Locked	NM	NM	NM	900	NA	NA	NA	NA	NA
	NM	NM	NM	450-500	NA	NA	NA	NA	NA

In general, the loads appear to be reasonable and consistent. However, it is difficult to extrapolate to a worst case in a Hanford tank, as even in the worst stuck condition, the vehicle should be able to free itself by removing material, even if slowly.

2.3.5 Retrieval and Decontamination

Retrieval and decontamination was conducted on Friday, 5/2. A commercial pressure washer delivering 40 gpm (nominal) at 3000 psi (nominal) was rented for the day, with an operator. The supply to the washer was from the local fire main through a filter. The decontamination rings always had at least one valve open to avoid dead-heading the pump. The pressure and flow are shown in figure 2.3-17. The pressure varied with number of spray rings operating, however flow was fairly constant at 34 gpm under all conditions. Based on this average flow rate and the total decontamination time of 89 minutes, the process used a total of 3026 gallons of water, which drained back into the test tank.

The vehicle was manually loaded with sticky hardpan and sludge, and allowed to dry for 2 hours prior to starting decontamination. The output of the pressure washer was connected to the decontamination system, and decontamination commenced by bringing the umbilical up through the spray area. Decontamination continued by raising the vehicle into the chamber, and directing the spray rings around to cover all areas. In addition, the wheels were powered up and rotated to expose all areas to the spray. Once the sluicer cleared the riser, it was also operated to expose all areas to the cleaning process, particularly the inside of the shroud. The spray wand was used to clean areas that the spray rings could not reach effectively, such as the interior of the motor mount boxes.

Photograph (30) shows the vehicle after being loaded with waste prior to retrieval. Retrieval is shown in photograph (31), the spray deluge during umbilical decontamination in photograph (32), and the spray deluge during vehicle decontamination in photograph (33). The rest of the retrieval sequence is shown in photographs (34) and (35), and the vehicle after decontamination in photograph (36).

Upon completion of the decontamination, the vehicle and umbilical were carefully examined, with the following results:

The umbilical, consisting of hydraulic hoses, water supply hoses, discharge hose, and electrical cables, was very clean. No material was observed on the outside of any of the umbilical components, however, some material was trapped inside the bunched umbilical components.
The high pressure of the spray, combined with the use of 0° jets resulted in damage to the hydraulic hose and water hose outer jackets. The damage was most likely caused when a single ring was operated and the pressure reached 300 psi. Lower pressures would probably be as effective and would not cause any damage. Conversely, 15° nozzles would reduce the impact of the spray and eliminate damage.
Centering the umbilical was somewhat difficult, and would not have been an issue with 15° spray nozzles. However this would not be an issue in an operational system because of the integral handling system design.
Raising and lowering the vehicle, and powering the wheels and cylinders was an effective means of exposing all areas of the vehicle to spray.
Visibility during spraying was very limited, and would probably be even worse if a video camera were used to monitor the process. With the test setup, the only way to stop the spraying was to secure the pump, which was not very convenient. An operational system should incorporate a bypass to allow shutdown of the spray for examination of the vehicle.
The wand proved to be very useful, however an improved spray pattern probably could eliminate the need for the wand.
The vehicle was examined after completion of decontamination, and was found to be very clean. The spaces between the lift arms and motor mount box, and the lift arms and motor casing were completely clean. Some residual material was noted in the front motor mount box and in the sluicer shroud. In both cases the dose traps could be eliminated by design. Some material also flushed out of the left rear wheel, however this material was forced into the space by the water jets during decontamination due the failure of the R.V. sealant. Again, this could be eliminated by design. The vehicle was not disassembled because of lack of time.
The addition of tracer dye to the simulants proved to be unnecessary, as the volume of water used for decontamination would have diluted the dye to the point of invisibility.
The high pressure spray conveyed material to the walls of the decontamination chamber and to areas of the vehicle and umbilical that had already been cleaned. A separate washdown ring for the chamber, multiple passes of washdown of the vehicle and umbilical, and perhaps additional spray rings would alleviate this problem. In addition, the spray ring nozzles should be changed from 0° to 15°, which would eliminate dead spots and reduce damage to the umbilical hoses and other soft materials.

3 ISSUES RESOLVED

3.1 Water Usage and Potential Tank Leakage

Description: Water usage is of importance because of the limited storage volume available for process water and waste removed from the Hanford Single Shell Tanks (SSTs). Constant use of raw water for the cleaning process will result in unacceptably high storage requirements or high cost to remove radioactive particles and dispose of the water safely.

Method: The volume and rate of water usage was determined by measurement during the testing. Average usage rates were calculated, and the feasibility of recirculating working water was demonstrated by the test setup. Standing water at 4 locations in the tank were measured, however these only reflected permeation through the simulated waste, not reflect transient levels.

Result: Total water volume needed to clean a hypothetical tank with about 6' of material was also calculated, and about 7.1 mgal, or about 5.5 times the volume of a 40' tank. This clearly points to the need to use recirculated water to clean an SST. Usage rates are consistent with recirculation of working water. Tank leakage could not be determined, quantitatively or qualitatively, since this would be highly dependant on the specific waste field and location of leaks in a tank. The ability to recycle water using medium pressure centrifugal pumps was amply demonstrated during the testing. The medium pressure water, 400 to 450 psi, was very effective at dislodging all but the hardest materials, and was very effective at mobilizing and conveying all the wastes.

3.2 Waste Removal Rate

Description: The waste removal rates for hard and soft materials are important issues as they will determine the length and cost of a tank cleanout. The removal rates are also important as they will drive the rest of the remediation process - transport to a Double Shell Tank (DST), the number of DST's required, and the rate at which material can be fed to the vitrification or other process.

Method: The waste removal rates were determined by measuring the waste field surface and computing the volume removed. Times were determined from the logged pressure and flow data. Determinations were made for hard and soft simulants.

Result: Waste removal rates varied considerably between hard and soft materials, as expected. The solids content by volume for both hard and soft materials were consistent with our experience at industrial sites and nuclear power plants. A total of 11.9% of the waste was removed during the testing. The measured rates were used to estimate the time required to clean a Hanford SST, and the results appear reasonable. Improvements in excavation of hard materials could be realized by equipping the vehicle wheels with a more aggressive tread, and by adding high pressure capability to the sluicer.

3.3 Mining Strategy

Description: Mining strategy is an important issue as it is closely related to removal rate, water use and the amount of water resident in the tank at any given time. Resident water needs to be minimized to reduce any potential leakage.

Method: The mining strategy was determined prior to the testing by logical analysis of the test process. Since the Free Water Scavenging Module (FWSM) was placed at a low point in the tank, the strategy was to clear a path to the FWSM by sluicing and excavating, and attempt to move waste and water along that path. Isolated pockets of waste and water that were created were cleared using the Vehicle Water Scavenging Module (VWSM) which was fed by the eductor on the sluicing head.

Result: The mining strategy was generally effective. Most of the waste and water were removed by the FWSN, except when it lost efficiency due to a pebble being stuck in the motive water nozzle. When the sluicer could be buried in water and waste, it, too, was an effective means of removing material. Mining strategy in a Hanford SST would be dictated by the specific conditions in the tank.

3.4 Operator Efficiency

Description: Operator efficiency is an issue related to the overall effectiveness of the cleaning operation, duration of the operation, and cost.

Method: Operator efficiency was evaluated by observation, and videotaped for later review. In addition, sluicer "on" time as a percentage of total time was calculated.

Result: It was quite clear that the most experienced operator had a better ratio of maneuvering to mining time, and did a better job of utilizing the sluicer and excavation capabilities of the vehicle. The most experienced operator utilized the sluicer 19% of the time against hard saltcake, while the least experienced had the sluicer on 62.4% of the time. Training and experience in operating the equipment will therefore have a significant impact on all aspects of an SST cleaning effort.

3.5 Deployment

Description: The ability to deploy the system without incident is a condition precedent to the actual cleaning operation.

Method: Deployment was demonstrated by deploying the vehicle and the FWSM in the tank. All operations were videotaped.

Result: Deployment of the vehicle and the VWSM were without incident, however the need to connect the sluicer discharge hose to the VWSM by remote control will have to be addressed in the future. The connection was made by hand during the test, and the connection components were standard cam-lock fittings. A fitting suitable for remote make and break will either have to be found or designed.

3.6 Mobility

Description: Mobility is required to be able to retrieve waste from a tank.

Method: The vehicle was driven around the tank prior to starting removal of material, to observe handling and agility. The vehicle was maneuvered extensively during the removal testing, including driving it into a hole in soft material. In addition, the vehicle was turned on its side to demonstrate the ability to recover from a knockdown. All activities were videotaped.

Result: The vehicle could be maneuvered to any location in the tank, even though the space was cramped relative to the size of the vehicle. The principal impediment to maneuvering was the umbilical jutting out directly behind the vehicle. This would be less important in a larger tank, or if the umbilical could be controlled with a winch, instead of manually, as was done during the testing. In addition, improvements in umbilical design and layout, and the means of attaching it to the vehicle would remedy this deficiency. When the vehicle was driven into a patch of softer material, it worked its way to the bottom of the tank, and sluiced and removed most of the material around it. The space was so confined that, in the time available, the vehicle did not free itself, although it continued to remove material. Given sufficient time a large enough area could have been opened up to permit maneuvering. The rollover test was successful, the vehicle was able to right itself merely by shifting its center of gravity with the lift cylinders and rotary joint.

3.7 Vehicle Recovery

Issue: Although it may not be necessary to remove the vehicle for maintenance during a cleaning, it must be possible to do so even if the vehicle is immobilized.

Method: The vehicle was positioned in the hole opposite the riser, and three tests conducted. The first was with the vehicle in free-wheel mode, the second with the wheels locked, and the third with the vehicle on its side and the wheels locked. The vehicle was pulled from the hole and cable angles and line tension measured. Loads on the riser and at the vehicle were calculated.

Result: The vehicle was easily pulled from the hole. Loads were similar for all three tests, primarily because most of the force was used to lift the vehicle. Loads in a Hanford SST would probably be higher unless a method were developed to reduce friction of the umbilical across the bottom edge of the riser. Note, however, that the vehicle drive system utilizes four fully independent drive motors, and this high degree of redundancy greatly minimizes the potential for an immobilizing failure.

3.8 Decontamination

Description: The vehicle and umbilical must be decontaminated if repairs are required, or if the equipment is to be re-used after cleaning an SST.

Method: A working decontamination chamber was fabricated for the testing. It was equipped with three spray rings, one with nozzles pointing down 45°, a second with the nozzles horizontal, and the third with the nozzles pointed up at 45°. Each spray ring had eight nozzles, and the entire ring assembly could be rotated +/- 22.5° to permit aiming water at all portions of the vehicle. In addition, the chamber was equipped with a manually operated wand. The vehicle was loaded with sticky sludge prior to starting decontamination. It was then decontaminated by raising it up through the riser, and moving it up and down through the spray rings, rotating the vehicle wheels, positioning the sluicer head and lift arms for effective cleaning, and using the wand to reach difficult spots.

Result: The umbilical and vehicle were both decontaminated effectively. Some residual material remained in between the hoses in the umbilical, where they were bundled with tie-wraps. Some residual material remained in one of the vehicle motor mount boxes, and in the sluicer shroud. This material could have been removed if it had been seen. In any event, dose traps such these would be eliminated by design, and the umbilical would be sheathed to eliminate dose traps. Note that the dye tracer mixed in with the waste was not useful in determining the presence of residue on the equipment. The presence of residue was easily determined visually. In a real environment, radiation monitors would be used to detect hot spots.

4 DISPOSITION OF TEST ITEM

The test mockup and apparatus was broken down the week after the test. All rental equipment was returned to the vendors. The vehicle, control station and other apparatus purchased or fabricated for the tests are stored at ARD Environmental, Inc. All original data, photographs, videotapes and the like are archived at ARD Environmental, Inc.

5 REFERENCES

ARD Technical Proposal: Single Shell Tank Waste Retrieval Alternative Retrieval Technology Demonstrations, 12/27/96.

Alternative Retrieval Technology Demonstrations Program - System Test Plan, ARD Environmental, Inc. 4/14/97.

Alternative Retrieval Technology Demonstrations Program - System Test Procedures, ARD Environmental, Inc. 4/21/97.

Lockheed-Martin Hanford Company Contract # MSH-SLB-A15120

Review of Prior SST Waste Retrieval Process Studies, WHC-SD-WM-ES-252, 9/27/93

Hanford Tank Waste Sluicing History, SD-WM-TI-302, 9/30/87

Decontamination System Study for the Tank Waste Retrieval System, EGG-WTD-11311 5/94

Initial ACTR Retrieval Technology Evaluation Test Material Recommendation, PNNL-11021

6 APPENDICES

Appendix 6.1 Figures

Figure 2.2-1.Test Site Layout