Generalities about USIM™ PAC
Introduction
The simulator
Use of USIM PAC
Preliminary plant design
Advanced plant design
Plant optimization and upgrade
Data processing
Conclusion
Introduction
USIM PAC is a powerful process simulation software package developed by BRGM
and commercialized since 1986. It is a user-friendly steady-state simulator that
allows mineral processing engineers and scientists to model plant operations
with available experimental data and determine optimal plant configuration that
meets production targets. The simulator can also assist plant designers with
sizing unit operations required to achieve given circuit objectives.
The
software package contains functions that can:
- manipulate experimental data,
- calculate coherent material balances, sizes and settings of
unit operations, physical properties of the processed materials,
- simulate plant operation and display results in tables and
graphs.
Widely used in industrial plant design and optimize, with more than one hundred
fifty licenses sold in thirty countries, this software has been continuously
improved, through successive versions, to make it more accurate and easier to
use.
These last years have seen significant developments in mineral processing
technologies, particularly in hydrometallurgy, bio-hydrometallurgy and mineral
liberation. In addition, it is now necessary to take into account the
environmental impact at each stage of a mining project, including water and
power consumption, waste treatment and disposal. USIM PAC 3.0 incorporates these
modern developments. Indeed, its structure and tools allow the user to take into
account, at the same time, a wide range of technological, economic and
environmental aspects.
The simulator
Process
modeling and simulation are used at all stages in the life of a mineral
processing plant: from process development to site rehabilitation, including
feasibility studies, engineering design, plant commissioning, plant operation
and upgrading right through to decommissioning. From the beginning, the
simulation-based approach gives an idea of the behavior and performance of the
future plant. This idea will be more and more precise owing to the
capitalization of knowledge acquired through laboratory tests, pilot plant
campaigns and plant operation. There is a continuous exchange between reality
and the virtual plant constituted by its steady-state simulator.
A simulator combines the following elements:
1) A flowsheet that describes the process in terms of successive unit
operations and material streams. This flowsheet encapsulates the experience of
the engineers responsible for the plant design or optimization. It can reflect
various scenarios so they can be compared against given criteria. It takes into
account numerous plant features such as reagents distribution, water recycling
or wastes treatment.
2) A phase model that describes the materials handled by the plant
(raw material, products, reagents, water, wastes) so that unit operations and
plant performance, product and reagents quality (grades and undesirable element
level), waste characterization (e.g. long-term behavior) for environmental
impact can be evaluated. The phase description is critical for analyzing and
optimizing the process. This statement reinforces the vital importance of field
data and sampling protocols.
3) A mathematical model for each unit operation. This model formalizes
the current scientific knowledge about the unit operation, and its level of
complexity depends on the data available and the targeted objectives (i.e.
flowsheeting, unit operation sizing or optimization). The model parameters -
dimensions, settings and calibration factors - are calculated or validated from
field data.
4) A set of algorithms for data reconciliation, model calibration,
unit operation sizing, full material balance calculation, power consumption and
capital cost calculation. These algorithms are interfaced to a set of data
representation tools.
As a result, the plant simulator constitutes a highly efficient communication
vector between the different actors who play a part in the plant life.
Use of USIM PAC
The simulation-based approach may be used for three main purposes.
The first one is the preliminary plant design.
The simulator is used to calculate an overall plant balance with quantitative
information on all streams (mass flowrates, grades, size distributions). This
yields selection of the key unit operations and a first estimate of investment
costs. At this stage, the simulator can be used to eliminate possible processing
routes.
The second one is the advanced plant design.
Based on detailed laboratory and pilot plant test data, the simulator is used to
predict the plant operation at a level sufficiently advanced for flowsheet
optimization and precise unit operation sizing. Used in advanced plant design,
the simulator is an invaluable aid at the plant commissioning stage or for
operator training.
The third classical application is plant optimization
(including audit, retrofit and upgrade), in which actual plant operation data is
used to build a simulation that mirrors the plant behavior. This approach is
classically used to improve the performance of operating plants.
Preliminary plant design
Figure shows the methodology used to produce a preliminary plant design. The
first step consists in assessing the plant requirements in terms of flowsheet
and stream descriptions based upon feed characteristics and main performance
objectives. A preliminary material balance is established by direct simulation,
which yields an ideal description of all the streams. The next step uses reverse
simulation to back-calculate the dimensions of the main pieces of equipment. The
final step simulates the future plant operation and calculates the capital
investment. This process allows the process engineer to compare several
flowsheets according to their technical performances and their financial
implications.
Advanced plant design
Figure illustrates the method for designing a new plant from a pilot plant
test. The first step uses material balance techniques to reconcile all
experimental data coming from sampling campaign during pilot plant test. The
second step consists in building a simulation of the pilot plant by calibrating
each unit operation model using these coherent data. After multiplication of all
streams by the scale factor, the next step uses reverse simulation to
back-calculate the dimensions of the main pieces of equipment in industrial
conditions. As previously, the final step simulates the future plant operations
in various configurations and calculates the capital investment.
Plant optimization and upgrade
Figure proposes a simplified box diagram of the methodology used for
optimizing the flowsheet of an existing plant (pre-control optimization or plant
upgrading due to new production objectives or operating constraints). The first
step uses material balance techniques to reconcile all available plant operation
data. The second step consists in building a simulation of the existing
processing plant by calibrating each unit operation model using the coherent
plant data. The final step is about using the simulator to test different
processing scenarios and analyze the simulation results in technical
(characteristics of the products, power drawn by the main equipment),
environmental (tailings stability, waste minimization, water recycling) and
economic (estimation of the capital cost investment and reactive consumption)
terms.
Data processing
Processing and analysis of field data require various algorithms that are
available in USIM PAC:
- Data reconciliation by statistical material balances gives a set of
coherent stream data from a set of experimental incoherent data coming from
different measurement sources (on-line sensors, pilot plant tests, laboratory
tests, etc.). Each data set is characterized by its own confidence level. The
only rules used in this algorithm are the material conservation laws.
- Direct simulation uses a sequential modular iterative algorithm. It
calculates all stream data from feed data and unit operation parameters. Direct
simulation can be used only on a selected sub-flowsheet or on a single unit
operation.
- Optimization algorithm seeks unit operation parameters that yield
output streams that match the objective streams as closely as possible. This
algorithm is used for unit sizing, model calibration or physical property
calculation, the difference appearing in the set of desired parameters:
dimensions, adjustment parameters or physical properties. The objective stream
can be an entire stream data set or some stream data parameters such as the d80,
a component grade or a recovery. It can also be an output unit operation
parameter such as the power consumption.
- Objective Driven Simulation (ODS) is a mix between direct simulation
and optimization. At each simulation iteration, it calculates, for each unit
operation, the parameters that meet a specified target. This algorithm is
generally used to improve the plant calibration previously done unit by unit
using optimization.
- Global optimization uses the same algorithm as the unit operation
optimization, however it applies to the entire flowsheet. It finds parameters of
specified unit operations that meet a plant performance objective. For example,
it can be used on a flotation circuit to evaluate the number and volume of cells
per bank that gives the best compromise between grade and recovery such that
profit is maximized.
- Capital cost estimation calculates the investment cost corresponding
to each unit operation with a cost model and the overall plant construction cost
with a hierarchical ratio system. The list and level of the ratios is entirely
configurable by the user.
- The supervisor can be used either as a sensitivity analysis tool or
for visual optimization. It calculates user defined plant parameters (global or
local performances, constrained parameters, etc.), named sensors, when some
input parameters (feed data or unit operation parameters), named actuators, are
varying. It is then possible to draw sensor variations as a function of
actuators and observe the sensitivity of a plant to given changes. In the case
of a multi-criterion optimization, it is easier to choose the best configuration
by examination of such a graph rather than by converting the target into an
objective function.
Each algorithm has many options for translating simulation objectives into
mathematical problems. However, these algorithms can be easily used with a set
of default options chosen for their ability to fit most situations. This is one
reason why USIM PAC can be used by process engineers as well as by researchers.
Conclusion
The simulator offers various powerful tools in response to
the increasing demand for a multi-criterion and global approach by plant
designers. It takes into account a wide spectrum of design criteria, including:
- Economic criteria such as capital cost, reagent and power
consumption, production quality in terms of valuable mineral grade or
undesirable elements level;
- Technical aspects with the evaluation of various
configurations and processing technologies, a complete and detailed description
of all material streams and their behavior during process;
- Environmental factors such as water consumption and
recycling, pollutant production or waste treatment.
USIM PAC is an extremely flexible simulator. It can be used
equally by process engineers for plant design or optimization, by researchers
for process development, as well as by academics for teaching process
engineering students.
The latest version, USIM PAC 3.0, represents a significant
milestone towards integrating different industries through a global approach. It
is currently possible to simulate treatment from the mine to the metallurgical
plant. Current studies on a global approach in urban waste management or metal
life cycle already use steady state process simulation techniques.