OPERATION AND PROCESS CONTROL DEVELOPMENT FOR A PILOT-

SCALE LEACHING AND SOLVENT EXTRACTION CIRCUIT RECOVERING

RARE EARTH ELEMENTS FROM COAL-BASED SOURCES
D. Addo, University of Kentucky, Lexington, KY
J. Werner, University of Kentucky, Lexington, KY
D, Threlkeld, Alliance Coal, USA
R, Bratton, Virginia Tech, Blacksburg, VA
R, Honaker, University of Kentucky, Lexington, KY

ABSTRACT protect domestic critical materials supply in the United

States, the U.S Department of Energy has funded
The US Department of Energy in 2010 has identified numerous research initiatives in the area of alternate REE
several rare earth elements as critical materials to enable supply, recovery, and extraction, in the United States.
clean technologies. As part of ongoing research in REEs Although researchers have reported the occurrence of
(rare earth elements) recovery from coal sources, the REEs in coal and coal byproducts, literature of recovery
University of Kentucky has designed, developed and is processes is limited for coal-based REEs. Zhang et al.,
demonstrating a ¼ ton/hour pilot-scale processing plant to 2015 [2] reported that the composition and complex
produce high-grade REEs from coal sources. Due to the distribution of REEs in coal and coal byproducts, as well
need to control critical variables (e.g. pH, tank level, etc.), as, the non-existent or lab scale REE recovery process for
process control is required. To ensure adequate process coal products is the reason for such limited literature. To
control, a study was conducted on leaching and solvent compensate, researchers have been exploring the existing
extraction control to evaluate the potential of achieving REEs enrichment processes and recovery techniques used
low-cost REE recovery in addition to developing a on other REEs sources. Typical REE extraction and
process control PLC system. The overall operational recovery methods include; roasting and leaching; ion
design and utilization of Six Sigma methodologies is exchange, precipitation, adsorption and solvent extraction
discussed. Further, the application of the controls design, (SX) [3-5]. The successful recovery of metals through
both procedural and electronic for the control of process acid leaching and solvent extraction has been noted in
variables such as pH is discussed. Variations in output literature. Research has also presented evidence on the
parameters were quantified as a function of time. Data unique extraction and recovery of REEs and other metals
trends show that the mean process variable was (e.g. copper, zinc) through acid leaching and solvent
maintained within prescribed limits. Future work for the extraction [2, 3, 6].
utilization of data analysis and integration for data-based Regardless of the exact REEs recovery process, the
decision-making will be discussed. flowsheet will involve a series of complex steps; hence,
the development of effective process control is vital to the
operation and recovery of critical rare earth elements. The
INTRODUCTION implementation of process control in chemical and
mineral-metallurgical recovery processes is critical to the
As part of their stewardship, Department of Energy
safety, operation, performance, productivity, quality, and
(DOE) has evaluated various rare earth elements critical
overall product recovery [7].
to the ongoing development and utilization of clean
Process control is a way of achieving both simple and
energy technologies. These elements are Yttrium (Y),
complex chemical industrial production goals using
Europium (Eu), Terbium (Tb), Neodymium (Nd), and
control systems in an economic and safe approach.
Dysprosium (Dy) [1]. Due to changes in green energy
Consequently, the benefits of process control
technologies and shifting markets on a global scale, the
future of the world’s energy market is uncertain. To

1
implementation in the metallurgical process industry have critical variables and effectively producing high-grade
been well researched. REEs. The pilot-scale leaching and solvent extraction rare
Ray et al., [8] presented a study where earth element recovery plant is designed to operate using
instrumentation and automation system control was both Allen Bradley PLC-based automated control system
utilized for a coal leaching pilot plant. The study explored and manual controls. The overall objective of this project
the use of automation (PLC) in increasing yields of is to design and develop a safe, efficient, and low-cost
chemical coal leaching processes. process control system with the capability of effectively
Wu et al., [9] studied model-based expert control operating a pilot plant producing high-grade rare earth
system which tracked the optimal pH of a continuous elements from coal-based sources based on six sigma
leaching process. Their research explored the use of a methodologies.
computer programmed expert control system that
combines both mathematical models and rule-based PLANT PROCESS DESCRIPTION
models.
In 2014, Dash et al [10] described how a PLC-based The schematic flowsheet diagram of the rare earth
control system was used to ensure the smooth operation element recovery process utilized at the University of
of a coal leaching pilot plant producing low ash coal Kentucky (UK) pilot plant is shown in Figure 1. The
product. The study applied supervisory control and data operation and control primarily focused on the leaching
acquisition (SCADA)-based system for a leaching pilot circuit feeding solvent extraction. The plant feed
plant treating Indian coal. The strong control system preparation consisted of feed size reduction via crushing
implemented at the pilot plant efficiently regulated plant and milling. As part of the leach circuit, an intermediate
parameters, improved data acquisition, and optimized acid make up tank is used to dilute and achieve the correct
plant operation. To achieve high-level process control of acid concentration for leaching feed. Acid leaching
the leaching process, the plant automation involved a consists of with cascading overflows with the pH
Rockwell-based PLC and other industrial instruments. controlled in the first feed tank. Tanks arranged in series
Overall research findings showed that the successful are used to achieve requisite residence time. Leaching
implementation of PLC-based automation and control occurs at 75℃ with pH controlled automatically by acid
system helped fulfill the objective of regulating process addition via PLC. The leached slurry is transferred into a
variables (e.g., chemical composition, temperature, and thickener. Here, the underflow solid is sent to the filter
pressure, etc.) and optimization. press for solids recovery and the clarified overflow liquid
There are other applications of process control PLC- together with the liquid recovered from the filter press is
systems, some are implemented in energy research, transferred to SX (solvent extraction) pre-treatment tanks.
manufacturing, education research, training, and This process consists of a reduction tank and
wastewater management control where measurement and neutralization tank prior to SX. In the reduction and
monitoring of pH and other process variables of a closed- neutralization tanks, ORP and pH of the solution
loop system in real time are critical [11-12]. respectively are manipulated for optimum SX feed by the
This paper will discuss the design and development same PLC system.
of a PLC control system with the capability of controlling

2
Figure 1. Process flow diagram of the pilot plant.

As part of wastewater treatment, the pH is adjusted and • The temperature control loop in Tanks 6 to 10 is
precipitate is removed by filter press. Also, the recovered implemented by a heater controller employing a
leach solids are washed and neutralized before disposal. standalone temperature control loop.
To assist in controls design, a piping and instrumentation
diagram (P&ID) detailing the leaching circuit has been
created detailing needed inputs and outputs (I/O) as Safety Considerations
shown in Figure 2 which shows the originally designed • The plant will have a Master ON/OFF switch. In
leaching circuit. Discussed below are leaching circuit the event of power failure; the plant must be
control loops and safety interlocks. restarted by the operator rather than
automatically resuming operation.
Control Loops • Agitators and pumps must be stopped whenever
• Feed flow control consists of thickener the tank liquid level is below impeller.
underflow, water makeup, raffinate recirculation, • Emergency stops (E-Stops) are strategically
and acid flow as shown in Figure 2. Control is located throughout the plant and are assigned
achieved by pump speed (rpm) adjustment using priority in the control system.
proportional-integral (PI) control. Tank level is • Safety checks or inspections must be performed
controlled by gravity overflow. before starting pumps, variable frequency drives
• pH and/or ORP control in tanks 6, 3, 4, and 19 as (VFD) or proportional valves.
shown in Figure 2 is achieved by reagent • Whenever a set point in pH, ORP, levels or coal
addition via proportional-integral-derivative feed is reached, pumps and valves should stop.
(PID) control.

3
Figure 2. Pilot plant leaching circuit P&ID (unscaled).

DEVELOPMENT OF PROCESS The overall controls design is based on four fundamental

principles; create smooth controls, controls with simple
CONTROL SYSTEM
interlocks, control the critical variables that matter, and
have non-competing control loops. In addition, critical
Control Methods
assumptions are made in the design of the control system
This section involves the investigation and
including
implementation of six-sigma methods to develop a PLC-
based automated control system. Benchmarking activities • Feed – the filter cake feeding system will be
and discussions were conducted to gather information on composed of 20-25% water and 70-75% solids.
critical processes in the pilot plant and variables critical to • Variability in feed material – feed rate will be
operation while exploring control theories and plans to controlled and monitored closely to maintain the
reinforce the development concept of the project. required percent solids feeding the leaching tank.
In the development of any process control system, the The system will be set to deliver 20 percent
control theory is vital. As part of the control theory, a solids.
select group of variables that are critical to the operation • pH Drift – the pH will drift up throughout the
of the leaching circuit was identified as well as previously circuit. Initial feed rate and recirculation rate will
used control steps in literature. be set constant with acid/base additions varying
• Percent solid concentration. according to pH set point.
• pH value. These assumptions provide guidance to the system design
process and control implementation.
• Eh or ORP.
An important phase of the control system design is
• Temperature.
the development of a control plan. The control plan is a
• Level and residence time. written description of how system control will be

4
achieved. It is also used for documentation and For example, solvent extraction is the downstream
communication. Subsequently, the control plan describes, customer of the leaching process. Figure 3 shows a
• What should be controlled? sample leaching control plan for REE recovery. Similarly,
• What should be measured and how often? a safety FMEA which evaluates the plant processes to
• What are the process specifications? identify potential areas of failure and relative impacts was
• How to ensure the process is performing well. developed in order to affect necessary change where
• What to do when it is not. needed (Figure 4). Finally, the total number of input and
The control plan is intended to be as a living document output (I/O) needed for the operation was determined as a
updated as control methods and measurements tools are result of the detailed control plan.
evaluated and improved. As part of a general rule to
establish an effective control plan for the leaching control,
the team utilized the basic understanding of the process
and available information to develop an initial control
plan. This included:
• Process flowsheets.
• Design/process failure mode and effects analysis
(FMEA).
• Lessons learned from literature.
• Team expertise or experience with the process.
Benefits of a control plan include reducing waste and
ensuring improvement in product quality, communicating
changes and documenting improvement, and meeting
downstream customer specifications.

Figure 3. Leaching process control plan.

5
Figure 4. Safety FMEA

PLC Control System Build

Several brands of PLCs were considered, but an
Allen Bradley (AB) PLC was deemed the most
appropriate choice for the pilot operation based on
support infrastructure available to the team.
Figure 5 shows the complete assembly of the plant
automation PLC systems with all I/O components. The
plant process control system build was based on Allen
Bradley 1769-L33ER CompactLogix PLC by Rockwell
Automation. Further components consisted of:
• 24VDC power supply mounted at the bottom
part of a waterproof enclosure.
• Ethernet switch miniature circuit breakers for the
controller.
• Ethernet adapter.
• 15 inch display.
• DeviceNet scanner module.
• 16 point DC input module.
• 8 point relay output.
• Analog I/O modules.
DeviceNet module will be utilized to communicate with
Figure 3. Pilot plant PLC system.
the various VFD drives. The main PLC control box
wiring layout is shown in Figure 6.

6
Figure 4. Plant PLC system wiring layout.

7
 After the build of the PLC system, small junction
boxes for connecting remotely located instrumentation
and equipment was built as shown in Figure 7. This
approach made wiring bundles and terminal connections
much easier. Each junction box is built primarily for a
pump control relay and to pass other sensor wiring to the
main PLC unit. The wiring layout for the remote junction
box is shown in Figure 8.

Figure 5. Remote junction boxes.

Figure 8. Remote junction box wiring schematic.

8
The plant automation process control is divided into four The PLC receives information from field input
levels. Figure 9 shows the architecture of the overall plant instrumentation, and sensors, processes the signal using
automation strategy. The communication of the SCADA the logic program and sends an appropriate response
and supervisory system with the PLC controller is signal to output devices.
possible through an Ethernet connection. Here the The system has RSLogix 5000 and FactoryTalk
SCADA system consists of a laptop computer that View. The RSLogix 5000 is the platform for ladder logic
handles analog and digital I/O’s through the PLC at the programming whiles FactoryTalk View is a human-
second level. SCADA in this control system is temporary machine interface (HMI) software application dedicated
and is as needed for testing purposes. In future work, the for machine-level operator interface devices (e.g.,
supervisory system level of control will incorporate Panelview Plus) by providing visualization of the system.
OSIsoft PI for data management and advanced statistical
process control. This data server will collect all field and
plant data to the SCADA system for advanced control and
optimization analysis.

Figure 9. The plant automation system architecture.

9
Instrumentation and Equipment
This section presents the different equipment and
instrument sensors selected for use in the pilot plant.
Selection criteria were based on low cost, suitability, and
reliability for industrial chemical purposes. For heating, a
400-watt fluoropolymer heater (Process Control HX
4229-P2) is used in combination with a conductive liquid
level probe (Process Control LC2H12) and thermocouple
(top mounted). For tank level control an ultrasonic analog
output sensor (AB 873P-D30AI-2500-D4) has been used
for solution holding tank or recirculation tanks. Two wire
submersion pH and ORP sensors (S8000CD & S8000CD-
ORP) assembled with 4-20mA module transmitter and
automatic temperature compensation (ATC) of Sensorex
make has been used for proportional-integral-derivative
(PID) based neutralization and reduction processes. Figure 10. HMI display design.
Masterflex peristaltic pumps (IP 7741-00 or LS 07528-
10) of external 4-20mA control capability using DB9
cable and variable area flow meters (FCA00151) proves
to be a good choice for plant flow applications.

Logic and GUI-HMI Design

A ladder logic program has been designed to operate
acid leaching tank pH controls in both manual and
automatic PID control mode. The same PID logic
approach is used for ORP control, wastewater pH
treatment, and level control.
The graphic user interface (GUI) HMI display is
available on a 15-inch color touch screen mounted on the
plant-based PLC controller box. Different display screens
are created to allow easy operator control of processes
whilst in the plant. The main screen of the HMI (Figure
10) presents the overview of the pilot plant leaching Figure 11. HMI operation monitoring display.
circuit as laid out in the P&ID. The next panel (Figure 11)
display presents the current date, time and the operational A control page push button (Figure 11) can be seen
control mode status for the critical variables for for each tank, which presents a new screen for controls.
monitoring purposes. The HMI also provides status The control page screen as shown in Figure 12, for
information about the set points, current variable reading, example, presents the graphical control panel for ORP
pump speed (rpm), and liquid levels. All the important and pH control system operation and each corresponding
data and trends are noted and stored in the system at a pumping system control buttons. When any pump images
regular time interval. are pushed or clicked, the operator is able to access
relevant trends corresponding to the individual loop
control. The design of the control system allows operation
in both manual control mode and automatic control mode.

10
 • pH control: The operator enters pre-determined
set point for the required acid leaching pH value.
Set point of the pH and the exact pH reading of
leaching material in Tank 6 are compared. The
acid solution was automatically introduced into
the leach tank 6 using a controlled peristaltic
pump. The pH control was achieved by the
manipulation of pH via proportional-integral (PI)
method. To reduce signal disturbance in the
readings by the pH electrode, a time average of
12 seconds was used to control the pH read.
After leaching, the slurry overflows into the thickener
where the solid underflow goes into a holding tank for
removal by filtration. Solid waste is treated and disposed
Figure 12. Variable PID control screen. of appropriately.

PLANT OPERATION SX Pretreatment operation

The thickener overflow and clarified leachate from
Installation and Shakedown the filtration process described previously are transferred
The leaching circuit for this project was assembled to tank 13 and tank 4 from the leaching process. In the
and installed at mine site located in Western KY. reduction tank, ORP of the solution is measured in
Leaching experimentation was achieved using Tanks 6, 7 millivolts and pH value of the solution is adjusted in tank
and thickener (tank 12) (see Figure 2). Each leaching 4.
vessel was filled with water and tested under proposed The valence state of iron in the SX feed is important
control conditions. Issues such as temperature lose, and to prevent SX contamination. The reductant solution was
leakages at connecting joints and pump tubing were automatically introduced into the mixing tank 3 using a
resolved. After successfully installing the system, controlled peristaltic pump. ORP control was achieved by
exploratory no-load testing of the leaching system was ORP manipulation via PID methodology. The pH of the
performed with only water to examine system neutralization vessel (tank 4) solution was maintained by
functionality and dependability before coal is fed into the PLC controller using the solution pH as the
system. controlling parameter. To reduce signal noise in the
readings by the pH electrode, a rolling average of 10
Leaching experimentation seconds was used to control the pH value.
Initial leaching experiments were conducted
following the shake down tests. It is important to note that Solvent Extraction (SX) operation
the initial leach tests were performed in a circuit modified The continuous SX circuit following leaching and
from that shown in Figure 2. The changes were to leach in pretreatment steps consists of loading, scrubbing, and
one concurrent stream utilizing leach Tank 6 and 7. This stripping stages. The SX mixer-settler unit is setup to
arrangement bypassed the first leaching leg of tanks 10, 1, internally recirculate the aqueous stream. The loading
2, 8, 9. The general circuit operational parameters are as stage is filled with leachate from the SX pretreatment step
follows: at a set constant pump speed (rpm) whiles stripping and
• Leach solution: Fill both 100-gallon volume tank scrubbing stages is filled with hydrochloric acid solution
6 and 7 with 350 liters of 1.2M sulfuric acid and at different concentration respectively at startup.
heat to Recirculation is needed to build the concentration of
• Temperature: Desired feed temperature of 75℃ desired REEs and reduce reagent consumption. Due to
is achieved by manually adjusting heater. low concentration of REEs in the feed stream, the SX
• Feed: Prepare coal to the required size, then circuit was operated for extended period of time to reach
convey over the top of Tank 6 through a screw equilibrium. Samples taken from solvent extraction
feeder system at a set feed rate of 50 lbs. per process are analyzed to ascertain the REE concentration.
hour while simultaneously feeding water at a Further treatment of stripped solution bleed was treated
controlled flow rate of 1 gallon per minute using with acid to selectively precipitate REEs to produce the
a rotameter. Agitators and pumps are kept in final rare earth oxide (REO) product.
running condition for proper mixing and to The trial-and-error method of PID control tuning
preventing clogging. used for pH, ORP, level and flow control was

11
implemented by RSLogix 5000 PID function block in the Table 1. Average of pH value recorded.
ladder logic programs. This function is used for level Upper Lower
control, control of ORP and pH in the leaching circuit and pH Mean Standard Control Control
also controlling the wastewater pH value. Tuning of the Setpoint pH Deviation Limit Limit
individual proportional, integral, and derivative term to (UCL) (LCL)
any particular system can be achieved using the trial and
error method for each process. In the PID function, the
2.00 1.96 0.0134 1.998 1.917
measured variable or process variable is compared to the
reference set point to calculate an error value which is
used to manipulate the controlled variable to maintain the
desired reference.

PROCESS MONITORING TRENDS

The HMI is also used as a monitoring tool for present
and past data of analog and digital types. These trend
graphs are designed and programmed using the Factory
Talk View SE software. Based on operators need; the
trend displays control performance at different times.
With different number of process variables being
maintained within specified limits in order for the plant to
operate smoothly, routine process monitoring will ensure
that system performance satisfies the operating objectives.
Statistical process control was evaluated to ensure Figure 14. A pH control trend during the leaching test.
that the process operates efficiently. Figure 13 and Figure
14 both show control charts showing the performance of Table 2. Mean pH values for the leaching test.
leaching experimentation pH for different case scenarios Upper Lower
at specific set points. In each case, the leaching pH Mean Standard Control Control
temperature condition was maintained at 75℃ and Setpoint pH Deviation Limit Limit
constant feed rate. (UCL) (LCL)
Table 1 and Table 2 show the test process data for Case 1
and 2 at pH set points 2.0 and 2.3 respectively.
2.30 2.40 0.054 2.565 2.245

By estimating the standard deviation (σ) of the

sample data, the control limits were computed to
determine the ± 3σ control limits to characterize the
system. Based on this result, the typical process variation
can be inferred to detect any anomalous or out of control
conditions. The small variability of data about the mean
value is attributed to common electric noise associated
with adjacent motor drives. Tuning of the PID will
consequently help improve small deviations about the
mean.
Figure 13. Leaching pH control chart.

12
 CONCLUSION AND 6. Kashiwakura, S., Kumagai, Y., Kubo, H., &
RECOMMENDATIONS Wagatsuma, K. (2013). Dissolution of rare earth
elements from coal fly ash particles in a dilute
The main goal for the development of a PLC-based H2SO4 solvent. Open Journal of Physical
process control system is to operate the pilot plant Chemistry, 3(02), 69.
leaching and solvent extraction circuit safely, accurately, 7. Bascur, O. (2019). Process Control and
and at the lowest cost for the recovery of coal-based rare Operational Intelligence. SME Mineral
earth elements (REEs). Processing and Extractive Metallurgy Handbook,
Literature review on leaching, solvent extraction, and 277.
process control assisted in a more complete understanding 8. Ray, T., Suresh, A., Dash, P. S., & Banerjee, P.
of mineral recovery processes and the different K. (2015). Instrumentation and automation
implementation methods of process control. The system at coal leaching pilot plant. International
implemented Six Sigma techniques were used Journal of Automation and Control, 9(2), 158-
successfully to scope the process and provided valuable 171.
knowledge and understanding towards safe design and 9. Wu, M., She, J. H., Nakano, M., & Gui, W.
development of process controls. (2002). Expert control and fault diagnosis of the
Based on the experimentation, testing logic leaching process in a zinc hydrometallurgy
programming, tuning and debugging the pilot-scale plant. Control Engineering Practice, 10(4), 433-
process control PLC system and logic described in this 442.
paper was able to operate and successfully control critical 10. Dash, P. S., Adari, P. R. K., Suresh, A., &
process variables (pH, ORP, level, and flow) in the Banerjee, P. K. (2014). SCADA-based control
recovery of REEs. system of pilot plant for chemical leaching of
As part of ongoing research, further system tuning is high ash Indian coals. International Journal of
essential to fully regulate process variables and fulfill the Automation and Control, 8(4), 362-380.
optimization objective. In addition, SCADA and 11. Alphonsus, E. R., & Abdullah, M. O. (2016). A
supervisory level data will be used to perform advanced review on the applications of programmable
process control and statistical analysis logic controllers (PLCs). Renewable and
Sustainable Energy Reviews, 60, 1185-1205.
12. He, J. Q., Chen, W., Gu, G. L., & Fu, X. R.
REFERENCES (2013). Automated Processing System of
Electrolytic Manganese Wastewater Treatment
1. Bauer, D., Diamond, D., Li, J., Sandalow, D., and Resource Utilization. In Applied Mechanics
Telleen, P., & Wanner, B. (2010). US and Materials (Vol. 271, pp. 204-207). Trans
Department of Energy critical materials strategy Tech Publications.
2. Zhang, W., Rezaee, M., Bhagavatula, A., Li, Y.,
Groppo, J., & Honaker, R. (2015). A review of
the occurrence and promising recovery methods
of rare earth elements from coal and coal by-
products. International Journal of Coal
Preparation and Utilization, 35(6), 295-330
3. Honaker, R. Q., Zhang, W., Yang, X., & Rezaee,
M. (2018). Conception of an integrated
flowsheet for rare earth elements recovery from
coal coarse refuse. Minerals Engineering, 122,
233-240.
4. Xie, F., Zhang, T. A., Dreisinger, D., & Doyle,
F. (2014). A critical review on solvent extraction
of rare earths from aqueous solutions. Minerals
Engineering, 56, 10-28
5. Zhang, W., & Honaker, R. Q. (2018). Rare earth
elements recovery using staged precipitation
from a leachate generated from coarse coal
refuse. International Journal of Coal Geology.

13