Martinka Mine No. 1, owned by Southern Ohio Coal Company, a subsidiary of Ohio Power Company and thus of American Electric Power Company, Inc., is a 2,500,000 ton per year (TPY) deep mine operating in the Lower Kittanning seam near Fairmont, West Virginia.
The mine operates nine continuous miner sections and two longwall sections, mining out approximately 280 acres per year. Due to the local topography and generally low cover overlying the mine, water infiltration into the mine is high in many areas. The mine discharges water at the rate of 2,000,000 gallons per day (GPD) on average and has peak discharges of approximately 3,000,000 GPD.
The amount of water being discharged has gradually increased as the mine continued its development. In 1984, mine studies indicated that an additional Acid Mine Drainage (AMD) treatment facility would be needed to augment the existing treatment facility and provide a total treatment capacity of approximately 6,500,000 GPD. This increase of 4,000,000 GPD would enable the mine to continue expansion of both its underground operation and its refuse area and to meet strict National Pollution Discharge Elimination System (NPDES) water quality discharge limits.
In an effort to meet its strict water quality discharge limits (See Exhibit I) at the new AMD facility, remove longwall emulsion oil from its discharge and be fully automated, mine personnel and American Electric Power Service Corporation Fuel Supply Department (AEPSC/FSD) personnel called for an AMD treatment facility that was both innovative in its treatment process and flexible in operation.
To satisfy these criteria, the mine and Barrett, Haentjens & Company and AEPSC/FSD conducted development work on the Turbojett Mixer. The Turbojett Mixer with no moving parts and easy maintenance, was effective in mixing a combination of liquids, reduced reagent costs, and provided flow flexibility. Based on these advantages the entire treatment process was designed around the Turbojett Mixer.
The Mixer also provided flexibility in operation through automatic controls that continuously monitor the quality of treated water and adjust the reagent feed rate. Short lag time between reagent addition and process control measurement assures accurate control.
Automatic control was essential to plant design since the plant location is several miles distant from the mine office and would have to operate most of the time unattended. The treatment system also had to be capable of directing water contaminated with longwall emulsion oil through the treatment process and, more important, be able to remove the oil before the water was discharged to the receiving stream.
Innovation and flexibility was also essential if the mine discharge was to meet the above limits established by the State of West Virginia, Department of Energy. Although the site was ideal from an underground collection and pumping standpoint, the surface location dictated discharging to Guyses Run, a high quality stream near Colfax, West Virginia.
Nearby residents expressed concern about the impact of the AMD facility on the stream. Southern Ohio Coal Company and AEPSC/FSD were able to meet the challenge of strict discharge limits by developing the Turbojett Mixer and by incorporating both it and other innovative features into the water treatment process.
A unique feature of the Guyses Run AMD Treatment Facility is the use of several water quality monitoring points, needed because of the high degree both of automation in controlling the treatment process, and the variability in the quality of the mine water entering the treatment facility.
Because of its high quality, 50% to 60% of the water at Martinka Mine No. 1, can be discharged directly to the receiving stream, bypassing the entire treatment process. When the water must be treated, it is usually because of high solids loading or contamination by oil emulsion; a water soluble oil mixed underground and used in the longwall hydraulic systems. When this oil is spilled or leaked underground, it can contaminate the entire mine water discharge for several hours.
Conventional AMD treatment techniques using hydrated lime were unsuccessful in treating this emulsion. To find a solution to this problem, Alden E. Stilson & Associates, Columbus, Ohio, were contracted by Martinka Mine to determine a method of treatment for the mine discharge contaminated with soluble oil No. 10, manufactured by Union 76.
| The laboratory study is summarized as follows: | |
|
A direct relationship was revealed between the turbidity of the oil/water emulsion and the concentration of the oil; | |
|
No correlation was observed between the oil concentration and the conductivity of the oil/water emulsion; | |
|
Lowering the pH of the oil/water mixture with a variety of acids would not break the emulsion; | |
|
Raising the pH of the oil/water mixture with a variety of bases would not break the emulsion | |
|
Treatment of the oil/water mixture with alum caused the emulsion to break. The results were a very slow-settling floc and complete removal of the haze from the water. Optimum dosage was about 25 mg/l (Al2(SO4)3o16H2O); | |
|
Treatment of the oil/water mixture with ferric chloride caused the emulsion to break. The results were a good settling floc and complete removal of the haze from the water. Optimum dosage was about 30 mg/l (as Fe Cl3 o6H2O); | |
|
The required dosages of both alum and ferric chloride were found to be independent of the oil concentration. Minimum dosages were found to be necessary to clarify the water. Increased dosages of either chemical above the optimum levels produced no beneficial effects on the results. These results were substantiated through further laboratory tests conducted by Sturm Environmental Services, Inc., Clarksburg, West Virginia. |
Based upon the above findings and field testing at the existing AMD Treatment Facility, a ferric chloride treatment system was incorporated into the overall design of the proposed Guyses Run AMD Treatment Facility.
The new AMD facility is approximately 4 miles from the mine site and will normally be unmanned, thus the need for automation. Based upon recommendations from outside consultants and previous AMD treatment experiences, it was decided that pH and turbidity would be the governing factors as to when the water would be treated or bypassed to the receiving stream. The operation and decisions for treatment and subsequent flow through the facility (Exhibit II) or bypass are as follows:
Water from the mine sump enters the monitoring building where the inflow pipe is tapped for two continuous sampling points; one for a surface scatter turbidimeter, and the other for a flow-through pH probe. If the water quality is below the turbidity set point and above the pH set point, the water is bypassed to a final polishing pond. If the water quality fails either of these tests, automatic valves switch the water flow to a raw water holding pond. The holding pond provides settling for solids that are occasionally encountered, and also provides additional holding capacity prior to treatment.
In the treatment building, the untreated mine water is resampled for turbidity. Should the turbidity be above the predetermined set point, the programmable controller turns on the ferric chloride feed system. The chemical is then automatically mixed with the water in an in-line mixer. The water then passes through the Turbojett Mixer where dilute caustic soda is added for pH adjustment. This is necessary even if the initial pH is within acceptable limits, because the ferric chloride acts as an acid and reduces the pH of the water. The caustic feed is controlled by a pH probe immediately downstream of the Turbojett Mixer in an open discharge channel. Since the Turbojett Mixer gives essentially instantaneous mixing, response time for pH control is very quick.
If required, flocculant can be added in the discharge channel to enhance settling of solids in the settling basin. The flocculant used is Dowell M-530 Anionic Polymer and is delivered to the discharge channel through an automatic pumping system.
As a final check in the treatment process, the clarified water from the Hydra System basins and the by-passed water are checked for pH and turbidity before entering the polishing pond for discharge. If the water quality is not within the set point parameters at this check point, it means a failure has occurred somewhere in the system. An alarm condition is annunciated at the mine security building and the treatment facility is shut down. This prevents any water of unacceptable discharge quality from entering the receiving stream.
The early development work on the Turbojett Mixer was sponsored by AEPSC/FSD to promote the development of new technology for treating mine water, with a particular interest in searching for new and better technologies to treat mine water at the Martinka No. 1 Mine, where a new AMD treatment facility would be constructed.
Early testing was done by treating water from the raw water holding ponds at the existing AMD treatment facility. Raw water was pumped directly to the 1.5 inch Turbojett Mixer, using a submersible pump. The unit was treating approximately 180 GPM which was directed back to the raw water pond. Samples for evaluation of the Turbojett Mixer discharge were collected in jars from the discharge spray.
Samplings of treated and untreated waters demonstrated that the Turbojett Mixer was capable of raising the pH of the water several pH units. These early tests consisted of a modified rockduster with a fluidized bed that was used to convey lime pneumatically to the Turbojett Mixer. Compressed air was used in various pressures to enhance the mixing process. Most of the testing done at this time was directed toward further development of the Turbojett Mixer itself. The unit experienced an unwanted build-up of powdered lime at the location where the pneumatically conveyed lime was introduced into the raw water stream.
Since parties involved with this testing were optimistic that these problems could eventually be resolved, it was decided in March, 1984, that testing of the Turbojett Mixer was to be continued in a location that permitted parallel operation with the existing AMD facility.
The purpose of this continued testing was to develop and test a Turbojett Mixer that could neutralize 800 GPM and not require any maintenance for at least 12 hours. During this test period, it was to be demonstrated that the pH could be manually adjusted to within 0.5 pH units of the 8.5 pH setpoint.
During subsequent testing with a 6" bore mixer, it was found that the lime build-up problem could be eliminated by providing for larger air flows through the unit. Forward air flow through the unit minimized the rearward splashing of the water jets within the mixer body. (See Exhibit III). This in turn alleviated the lime build-up problem. It was found, however, that as water flowrate through the unit was increased (by increasing hole size), the induced air flow diminished. This, in turn, caused unacceptable splashing at the rear of the unit. Finally, an 8" bore Turbojett Mixer was built to permit treatment of 800 GPM water flow with sufficient air flows (in the range of 400 CFM), to alleviate any lime build-up problems.
At this point in the testing/development program, AEPSC/FSD and mine personnel decided to use sodium hydroxide as the neutralizing reagent for the new Guyses Run AMD Treatment Facility. Turbojett Mixer development using hydrated lime was suspended, and testing with sodium hydroxide was begun.
Since low flowrates of sodium hydroxide needed to be introduced into the flow stream, it was decided to introduce the reagent flow at one point at the rear center of the Turbojett Mixer, rather than use a separate set of injection nozzles connected to a separate chamber in the Turbojett Mixer. The single nozzle used in both the testing and the final installation is a hollow cone spray type. At 40 psi, this nozzle will atomize the caustic solution to particles with a mean diameter of 750 microns.
To test this arrangement, caustic-water solutions were mixed to various strengths to achieve the desired end point pH. This arrangement simulated the effect that was to be used in the final installation. Field testing indicated that this was a viable arrangement to treat the water from the deep mine.
The chemical feed arrangement used in the final installation of the four Turbojett Mixers at the Guyses Run Treatment Facility was generally that previously tested, except for a few adaptations.
Using design suggestions from Craft Products in Pittsburgh, Pennsylvania, the sodium hydroxide reagent stream (normally 1.44 gallons per hour) is diluted with a much larger flow stream of water (2 GPM). This dilution allowed a larger injection nozzle size, which is less likely to clog, while maintaining a steady atomizing spray effect as the flow rate of reagent varies. other benefits of this design are reduced personnel hazards because of the diminished reagent concentration in the stream being sprayed into the rear of the mixer. Also, to reduce any possible pulsation of reagent feed resulting from individual strokes of the piston feed pump, a separate retention/mixing chamber of several gallon capacity was placed in line with the diluted caustic feed piping. The material construction of the Turbojett Mixers is carbon steel covered with 80 mil Plastisol coating, inside and out. The bore piece containing the injection holes was machined from PVC tubing. The mixers are easily disassembled, using one single latch clamp.
In an independent test, conducted by another coal company, it was found that the Turbojett Mixer was effective in reducing the suspended iron content of the discharge stream leading from the treatment facility.
In this test, a Turbojett Mixer with a 4" bore was operated with an upstream pressure of 30 psi, and a flowrate of 250 GPM. During this test, the discharge of the Turbojett Mixer was sprayed into a receiving pond (Pond #1). This receiving pond, in turn, flowed into a second pond which discharged to a stream. Assuming the ponds did not short circuit, pond #1 water should be fully replaced in four days, and pond #2 water would be replaced in fifteen days.
During testing, samples were taken from the raw water, from the Turbojett Mixer discharge, from pond #1 discharge, and from pond #2 discharge. During the first four days of testing, samples were taken three times daily and then eight more times during the subsequent twelve days.
The average of the laboratory results indicated that the dissolved oxygen level rose from a level of 2.3 ppm in the raw water to a value of 6.9 ppm immediately downstream of the nozzle. After the pond water had turned over, the dissolved oxygen level at the pond discharges were 9.7 ppm and 10.2 ppm for ponds #1 and #2, respectively. The pH of the raw water was measured at an average value of 6.8 and the pond #2 discharge had a value of 8.1.
Likewise, the ferrous iron concentration was measured at an average value of 19.2 ppm for the raw water and was reduced to a value of 4.3 ppm immediately downstream of the Turbojett Mixer.
The sludge removal system selected for this installation is the "HYDRA" System, as manufactured by Barrett, Haentjens & Company. This system has operated successfully in several other AMD treatment facilities as well as the existing treatment facility at the Martinka No. 1 Mine. The system has no parts moving through the sludge bed to resuspend fine settled material. The sludge removal system is fully automatic, using separate input times to control the duration of sludge pumping from each section of the settling basins. Another time clock controls how long the system is in operation each day.
Basically, the "HYDRA" System is an extended, segmented, pump suction pipeline. The system rests on the bottom of the settling area and allows solids to be withdrawn from where they have settled (Exhibit IV).
The pump suction network is partitioned into segments through the use of automatic control valves. Each segment consists of a collection of twenty, 0.75 inch inlet orifices which gather settled solids. The orifices are evenly spaced to produce a uniform sludge withdrawal pattern throughout each segment (Exhibit V). An automatic control valve isolates each segment from the pump suction. These control valves open and close in sequence to distribute the pump suction to selected segments. The control valves are pneumatically actuated by an automatic control panel which sends out timed, sequential signals. The duration of each valve's open time (i.e., how long pump suction is directed to a particular area) is individually programmable.
The prime mover in this general arrangement is the slurry PUMP. It draws the sludge bed through the suction orifices, suction piping, control valves, and ultimately to the pump suction. From here the sludge is pumped back into an abandoned section of the mine. Successful operation of this system does require that the sludge (liquid-solids mixture) be in a flowable state.
The "HYDRA" system suction grid is evenly spread over the bottom area of each of two basins, 30 feet wide by 120 feet long. Each basin contains twenty, 2-inch Cla-Val globe-type, air actuated control valves. During normal operation of the system, the control panel sends an air signal to actuate two valves at one time. Each valve passes approximately 100 GPM of sludge.
All valves are connected to a common 4-inch diameter PVC header. The header passes through the tank wall and connects to the pump suction. The pump utilized is a Hazleton 4" D type CTE horizontal centrifugal slurry pump.
The primary control system for the plant uses eight Allen-Bradley PLC-4 Microtrol programmable controllers. Controllers are tied together in a 4,000 foot loop. This arrangement allowed the controllers, located in four places, to be connected to their respective inputs and outputs with a minimum amount of control wiring. Six of the controllers are in communication with each other via twin axial cable, a single pair of shielded wires. The controllers sense "on-off" signals at the controller inputs. The controller logic produces appropriate "on-off" outputs with necessary timing delays. All controllers in the control loop may be monitored or reprogrammed at any controller in the loop.
For additional diagnostic aids, the input and output terminal strip of each controller is outfitted with indicator lights to show the status of all inputs and outputs. The controller memory has a back-up battery power to allow the programmer to retain its memory during power outages. The system will automatically restart on power resumption.
Various alarm and emergency signals passing through the programmable control system are transmitted over phone lines through a communication system manufactured by QEI. This arrangement allows personnel on duty continuously at the mine site to monitor the operation of the remotely located AMD plant and respond to plant faults or personal injury as necessary.
The programmable controls allow modification to the treatment system's operating sequence, as deemed necessary by changing water quality conditions; Substitute, or back-up program instructions are stored on electrically erasable, programmable, read only, memory chips (EEPROM). These chips can be used to store back-up programs, or can act as a convenient medium to transmit program information between chip supplier and customer.
A multi-track recorder is used to record the values of the treated water pH and turbidity. The pH and turbidity of water entering the treated water pond is also recorded along with the flowrates of the treated water and bypass streams. The chart recorder prints appropriate scales for pH, turbidity and GPM on the chart paper, along with the time and date of recording. This permanent record output provides a detailed summary of the plant operation.
The Guyses Run Treatment Facility was declared commercial on January 1, 1986, and except for the usual start-up problems, has lived up to early expectations of being innovative and flexible.
Early problems that have been identified are in the ferric chloride delivery system, where it is essential that the product be handled according to the manufacturer's recommendation.
Although several leaks in the delivery system have shown how corrosive ferric chloride can be, mine personnel are pleased with the product's effectiveness in treating the oil emulsion.
Initial operating information has shown that the neutralizing agent, caustic soda, will adhere to the inside of the Turbojett Mixer and cause a build-up, though this is slight and easily removed by the use of handscrapers when the system is not operating. It is expected that the build-up will need removal four to five times a year.
Experience has also shown that the Turbojett Mixer should be designed to pivot into a vertical position to facilitate reassembly.
Air flow through the Turbojett Mixer during operation has been good, with approximately 400 CFM being induced, an indication of its aeration capability. And, prior testing indicated that these air flow values may be further enhanced.
Subsequent testing by the authors showed that discharge tube extensions added to the Turbojett Mixer were able to induce more air flow through the unit. Air flow measurements were taken for the unit as installed, and with five extension devices adding up to 18 inches to the discharge tube length.
In adding these extensions, air flow was increased from 310 CFM to 690 CFM. Add-on extension devices are now being added to the basic design.
Based on the successful operation of the Guyses Run Treatment Plant, AEPSC/FSD has decided to include these innovative and flexible designs into future treatment plants. Treatment facilities currently being designed for the Meigs Division include the use of Turbojett Mixers and a Hydra Sludge Removal System.
EXHIBIT I
|
DISCHARGE |
30-DAY |
DAILY |
|
PARAMETERS |
AVERAGE |
MINIMUM |
|
Fe |
1.5 |
1.5 |
|
Mn |
1.0 |
1.0 |
|
TSS |
35 |
70 |
|
PH |
(BETWEEN 6 AND 9 AT ALL TIMES) |
|




