Title: Membrane processes in environmental technology
1Membrane processes in environmental technology
- Introduction to the
- Practical Exercises
- Hans Saveyn Prof. Paul Van der Meeren
23 Practical exercises
- Reverse Osmosis of saline water
- Ultrafiltration of nanosuspensions
- Membrane performance and cost evaluation with the
simulation program Memsys
3General info
- Practical exercises presence obligatory
- Team work (rotating groups of 3-4 people)
- Every week 1 exercise/group
- One report has to be made per group and handed in
before the start of the next exercise - Preparation is necessary
- Quotation has important impact on final mark (for
part of prof. Van der Meeren) !
41.Reverse osmosis of saline water
- Aim desalinate NaCl solution by RO
- Practical work on a pilot scale RO unit
- Calculate specific parameters for the membrane
and operation
52.Ultrafiltration of nanosuspensions
- Aim filter nanosupsenions of silica by crossflow
ultrafiltration - Practical work on a labscale filter
- Determine several parameters and investigate the
effect of concentration polarization
63.Membrane performance and cost evaluation
- Aim to use a membrane simulation program
(Memsys) in order to evaluate membrane
performance and cost of operation - Investigate the influence of particle size and
concentration on steady state cake thickness and
steady state permeate flux - Calculate the time averaged permeate flux and
operation time needed for the different scenarios - Evaluate the total treatment cost of water by UF
membrane technology
73.Membrane performance and cost evaluation
- ProgramMemsys (by Sandeep Sethi and Mark R.
Wiesner) - Computer Simulation Model for Performance and
Cost Modeling of Ultrafiltration and
Microfiltration - References
- Sethi S. (1994) Performance and Cost Modeling of
Low-Pressure Membrane Filtration Processes. M.S.
Thesis, Rice University, Houston, Texas. - Sethi S and Wiesner MR (1995) Performance and
Cost Modeling of Ultrafiltration. Journal of
Environmental Engineering, 121 (12) 874-883 - Program freely downloadable from
http//www.ruf.rice.edu/wiesner/memsys.htmlDOWNL
OAD
8Memsys simulation program introduction
- Operates in MS-DOS mode (runs also on Windows
based PCs)
9Memsys simulation program introduction
- Easy operation with menu structure
10Memsys simulation program introduction
- MEMSYS is a computational tool for evaluating the
performance and cost of crossflow ultrafiltration
(UF) and microfiltration (MF) systems. - The computer program is written in FORTRAN and
runs on the PC-DOS environment. It includes a
menu-driven interface for running model
components, manipulating data, and graphically
analyzing performance results. - A performance module estimates the steady-state
permeate flux and initial removal of particles
and organic matter. Capital and operating costs
can be calculated subsequently, using the
permeate flux estimates provided by the
performance model. Alternatively, the user may
run the cost module independently, and input
permeate flux information directly to evaluate
costs. - The variation of permeate flux, cake thickness,
and other dependent parameters, along the length
of the membrane, can be plotted and analyzed
using the graphics module. -
11Memsys simulation program introduction
-
- Synopsis of model The model estimates
steady-state permeate flux based on particle
transport by - shear-induced diffusion
- Brownian diffusion
- permeate drag
- bulk convective transport of the boundary layer
along the membrane surface - Initial removal of particles and macromolecular
natural organic matter (NOM) by the membrane is
estimated using simple expressions for particle
capture in an idealized membrane pore. These
expressions are based on theoretical aspects of
hindered solute transport in membrane pores.
12Memsys simulation program introduction
-
- Macromolecular dimensions are calculated from
molecular weight using a power law expression. - For the purposes of calculating permeate flux, a
scheme for weighting particle and macromolecular
sizes based on their influent concentrations and
relative diffusivities is used to reduce the
heterodisperse suspensions input to the program
to an average particle diameter.
13Memsys Limitations and simplifications
- The model holds in situations where the solute
accumulation near the membrane is thin relative
to the radius of the membrane element. The
assumption of constant bulk suspension
concentration limits the model to situations
where the thickness of the cake formed on the
membrane is small compared to the element radius.
The solution would thus be an approximate one in
situations where the channel gets plugged. - The performance model holds only under laminar
flow of the suspension through the membrane
modules. - Polydisperse suspensions are considered by
calculating an average parameter or mean particle
size.
14Memsys Limitations and simplifications
-
- Transient effects are not considered in
evaluating the performance of the membrane
filter. In other words, long term phenomena, such
as pore fouling, are neglected. - Gravitational effects are not considered. Hence,
all solute particles are assumed to be neutrally
buoyant. - The model to calculate permeate quality assumes
membrane pores to be cylindrical and the solute
particles to be spherical. Solute-Membrane
interactions and concentration polarization are
not considered.
15Memsys Main Menu
16Memsys Main Menu
- Global options
- Set options on how to calculate certain
parameters - Data Manager
- Enter data concerning operation, feed composition
and costs - Performance Analyzer (PA)
- Evaluate operation (permeate flux and quality)
based on entered data from the Data Manager - Cost Estimator (CE)
- Evaluate process cost based on permeate flux
data, operating time (to be calculated manually)
and entered data from the Data Manager - Results Display
- Shows earlier calculated PA or CE data
- Graph View
- Allows plotting of certain parameters
- Print
- Dos shell
17Behind the program
18Memsys Model development
- Cake build up on membrane depends on total
diffusivity - Total diffusivity is composed of Brownian
diffusivity Dbrn and shear-induced hydrodynamic
diffusion Dsh - Brownian diffusivity
- Shear induced diffusion (particle motion under
shear flow)
19Memsys Performance analysis
- Small particles (lt0.01 µm) exhibit high Brownian
diffusivity (random walk) - Larger particles (gt1 µm) exhibit high
shear-induced diffusivity (are easily dragged
away) - Combination of both diffusivities allows to
predict the steady state cake thickness and
steady state permeate flux by an iterative method
for a given membrane system and feed stream
composition - This is done in the software by running the
Performance Analyzer, after having entered the
membrane and feed data in the Data Manager
20Memsys Operating time and time averaged permeate
flux calculation
- After calculation of the steady state permeate
flux, and other cake parameters, the operating
time can be calculated from the recovery
requirements, the permeate flux evolution and the
steady state permeate flux - The time averaged flux can be calculated from the
permeate flux evolution and the steady state
permeate flux - Both parameters are needed for the cost
estimation
21Memsys Operating time and time averaged permeate
flux calculation
- Working principle of membrane set-upWasting of
the bulk liquid is only supposed to happen during
backflushing (continuous recirculation of bulk
liquid during filtration)
22Memsys Operating time and time averaged permeate
flux calculation
- Recovery expressions
- Reorganisation to get operating time t0
- Time averaged permeate flux Vw average needed !!!
23Memsys Operating time and time averaged permeate
flux calculation
- To calculate the time averaged permeate flux, the
permeate flux evolution as a function of time has
to be considered - The permeate flux will decline in time until the
steady state permeate flux (as given by the
Memsys program) is reached at time tss (steady
state time) - From tss onwards, the flux will remain constant
Permeate flux
time
tss
24Memsys Operating time and time averaged permeate
flux calculation
- Two possible scenarios
- Operating time t0 lt tss
- Operating time t0 ? tss
Permeate flux
time
t0
tss
Permeate flux
time
t0
tss
25Memsys Operating time and time averaged permeate
flux calculation
- Permeate flux evolution as a function of time
before steady state is reached
26Memsys Operating time and time averaged permeate
flux calculation
- Calculation of steady state time tss t for
which Vw(t)Vw steady state
27Memsys Operating time and time averaged permeate
flux calculation
28Memsys Operating time and time averaged permeate
flux calculation
- Final calculation of t0 working algorithm
- First calculate tss
- Then calculate t0 with the first expression
- If calculated t0 lt tss, then keep this t0 value,
otherwise calculate t0 again by the second
expression
29Memsys Operating time and time averaged permeate
flux calculation
- Final calculation of Vw average
30Memsys Operating time and time averaged permeate
flux calculation
- Parameters necessary for calculation of the time
averaged permeate flux Vw average and the
operating time t0 are partly input parameters
entered in the Data Manager (Rm, Rrec, Vbf, tbf)
or can be retrieved from the simulation program
after running the Performance Analyzer (?b,Vw
clean water,Rst, Vw steady state) - After calculation of the time averaged permeate
flux and the operating time , these values have
to be entered in the program - Time averaged permeate fluxGlobal options?
Permeate Flux?User defined flux - Operating timeData Manager?Cost data? Item 5
Operating Period
31Memsys Cost Estimation
- When having entered the correct time averaged
permeate flux and operating time, the Cost
Estimator can be run to produce a cost estimation
of the process
32Exercise
- Evaluate the steady state permeate flux and
steady state cake thickness for the default data
series - Calculate the values of ?c, tss, the operating
time t0 and the time averaged permeate flux Vw
average - Evaluate the total treatment cost (/m³) of the
process by entering the calculated time averaged
permeate flux and the operating time in the
program and then running the Cost Estimator - Be careful when converting non SI - units !!!