You are on page 1of 4

Water Management

Z. ADI PUTRA, TMC Chemical BV, Eindhoven,


The Netherlands

Improve performance and reduce costs of


wastewater treatment plants
The increasing costs of a wastewater plant by reducing the wastewater load to Proposed modifications to system.
treatment plant (WWTP) motivated a the collective WWTP. A proposed modification from plant per-
major specialty chemical plant to improve The schematic of the internal treat- sonnel is shown in FIG. 3. The condensed
the performance of its treatment unit. ment unit is shown in FIG. 1. The wastewa- water (2,000 kg/hr) from the first effect
It was determined that the wastewater ter contains roughly 12 wt% salt, 80 wt% of the multiple-effect evaporator has the
load to the collective WWTP needed to water and some organics. From the pit, lowest concentration of total organic car-
be reduced. A water pinch analysis was the wastewater is pumped to a multiple- bon (TOC). This water can be directly re-
performed to determine the minimum effect evaporator to evaporate the water. cycled and combined with the freshwater
freshwater consumption and wastewater The concentrated wastewater is then sent stream, which will reduce the freshwater
discharge. Then, a linear programming to a centrifuge unit to remove the salt. The consumption, as well as the wastewa-
model was created to develop the opti- condensed water is collected and sent to ter discharge, by 2,000 kg/hr. The esti-
mum water network required to reach the a centralized wastewater treatment plant mated total cost of the modified system
target. The result shows improved savings in the complex. Two other wastewater is €45,000/yr—a reduction of 42% from
and a practical water network solution. streams are considered relatively clean (in the cost of the existing system.
other words, negligible amounts of salt) The proposal prompts a number of
Wastewater treatment system op- and sent directly to the collective WWTP. questions: Is the proposed modification
eration. A major specialty chemical The various flowrates of the current acceptable technically? If not techni-
plant uses a significant amount of water setup are illustrated in the water network cally acceptable, what is the acceptable
for its production, cleaning and washing. in FIG. 2. Freshwater consumption is 4,650 solution, and how much savings can be
Those wastewater streams are treated in kg/hr. Most of this water is used for wash- achieved? Is there a better solution?
an internal wastewater treatment unit, as ing; some reactions in the plant produce To answer these questions, a systemat-
well as in a collective WWTP in the com- additional water. Therefore, the total ic methodology called water pinch analy-
plex. The increasing cost of the collective wastewater discharge is 4,950 kg/hr. The sis is performed. This method starts with
WWTP motivated plant personnel to im- total cost of the water system is approxi- gathering data on what contaminants are
prove the performance of the treatment mately €77,000/yr. considered, and in what concentration
Existing situation
Collection Collection
To WWTP vessel 1 vessel 2
Unit 1

Unit 2 Hot oil out Hot oil in

Plant
Fresh water
To WWTP
1st 2nd 3rd
Unit 3 effect effect effect

CW out CW in
Unit 4
Fresh water

Unit 5 Pit
Centrifuge Salt
(Unit 6)
Concentrated organics

FIG. 1. Schematic of the existing operation of the multiple-effect evaporator.

Hydrocarbon Processing | DECEMBER 2015 75


Water Management

level. Based on this data, the proposed lumped into a single contaminant, and composite curves. Once the target is de-
modification can be evaluated. As shown wastewater target flow for recycle and fined, the water network can be optimized
later, the considered contaminants can be discharge can be obtained via source-sink to ascertain if any stream can be reused or
recycled. Further explanation on water
TABLE 1. Source streams pinch analysis and source-sink composite
curves can be found in literature.1, 2, 3
Number Flow, kg/hr Concentration, ppm Load, mg/hr Stream name
After the optimum water network is de-
1 950 0 0 Fresh veloped, the reused or recycled streams are
2 2,000 50 100,000 Evaporator 1 combined with the internal plant’s Excel-
3 1,900 500 950,000 Evaporator 2 based mass balance to check if the solution
4 1,000 600 600,000 Evaporator 3 has any effect on the product quality.
5 50 1,000 50,000 Dryer
The produced wastewater stream from
the plant with the reused or recycled
Current water network stream(s) will then be sent to a proprie-
4,650 kg/hr 650 2,000 Collection tary evaporator model to check if the con-
Fresh water Unit 1 Evaporator 1 vessel 1 centration of the stream(s) to be reused
Water cost (thousand €/yr): 37 or recycled will be affected. The overview
650 1,900
Unit 2 Evaporator 2 of this integrated model (Excel plant mass
Collection 4,950 kg/hr WWTP
vessel 2 balance, proprietary evaporator and pinch
50 1,000 Wastewater cost
Unit 3 Evaporator 3 (thousand €/yr): 40 analysis models) is shown in FIG. 4.
Total cost (thousand €/yr): 77
1,300 50 Data collection. Data collection was
Unit 4 Dryer
performed via intense communication
1,200 with plant personnel. Some of the lines
Unit 5
lack flowmeters, so their flowrates are es-
800 timated based on mass balance.
Unit 6 To check if the water reuse/recycle is
acceptable, the chemicals involved can be
FIG. 2. Schematic of the existing water network.
lumped into a common property: TOC.
Proposed design This lumping approach simplifies the
2,650 kg/hr 650 2,000 Collection problem and makes water pinch analysis
Fresh water Unit 1 Evaporator 1 vessel 1 via source-sink composite curves possible.
Water cost (thousand €/yr): 21 Lists of available water streams (source
Combined water (ppm): 22 650 1,900
Unit 2 Evaporator 2 542 ppm streams), required water streams (sink
Collection 2,950 kg/hr WWTP
vessel 2
streams) and their corresponding concen-
50 1,000 Wastewater cost
Unit 3 Evaporator 3 (thousand €/yr): 24 trations are listed in TABLE 1 and TABLE 2.
Total cost (thousand €/yr): 45
1,300
Unit 4 Dryer
50 Reduction: 42% Analysis of proposed modification.
Based on the collected data, an analysis
1,200 was performed on the proposed modifica-
Unit 5
tion. The condensed water from the first
800 effect evaporator (Evaporator 1) is indeed
Unit 6 the “purest” stream, and it has a TOC of
approximately 50 ppm. However, Unit
FIG. 3. Proposed modification to the wastewater treatment system.
1 and Unit 2 have stricter water require-
Condensed water from Evaporator 2 ments of 10 ppm and 20 ppm, respective-
ly. Therefore, the proposed modification
Condensed water from Evaporator 1 of combining the condensed water with
the freshwater stream cannot be justified.
To keep the required concentrations
in Unit 1 and Unit 2, only 1,800 kg/hr
Plant of water can be recycled instead of the
Freshwater Multiple-effect Wastewater
proposed 2,000 kg/hr, resulting in a cost
Excel mass- evaporator model
balance model reduction of only 38%. This scenario is
illustrated in FIG. 5, where the freshwater
intake is 2,850 kg/hr and the wastewater
discharge is 3,150 kg/hr.
This short analysis answers the ques-
FIG. 4. Interaction between the calculation models.
tion of whether or not the proposed modi-
76 DECEMBER 2015 | HydrocarbonProcessing.com
Water Management

fication is acceptable. To investigate if a • Higher wastewater discharge To solve this issue, the “fresh” waste-
better solution is available, the source-sink • Higher energy required for the water stream directly from the pit is used
composite curves are made (FIG. 6). The evaporator. to wash the centrifuge. This stream is rela-
target for the freshwater intake is far below
Modified proposed design
the present consumption, which is 910 kg/ 1,800
hr. The wastewater discharge is also re-
2,850 kg/hr 650 2,000 Collection
duced by 79% to 1,210 kg/hr. Fresh water Unit 1 Evaporator 1 vessel 1
As outlined by the source-sink com- Water cost (thousand €/yr): 23
650 1,900 200
posite curves, these targets are used as the Combined water (ppm): 20 Unit 2 Evaporator 2 511 ppm
benchmark prior to any water network de- Collection 3,150 kg/hr WWTP
vessel 2
sign. To determine how the water network 50
Unit 3 Evaporator 3
1,000 Wastewater cost
(thousand €/yr): 25
should be designed, a superstructure of the Total cost (thousand €/yr): 48
sources and sinks is developed considering 1,300 50 Reduction: 38%
Unit 4 Dryer
all possible connections between all water
sources and all water sinks. 1,200
Based on this superstructure, mathemat- Unit 5
ical formulations are developed and solved 800
to minimize freshwater consumption (i.e., Unit 6
total cost involved). Details on how the su-
perstructure is built and the mathematical FIG. 5. Proposed design modified due to contaminant constraint.
programming are explained in literature.1, 2, 3
1,600,000
One of the optimum water networks Wastewater discharge
is shown in FIG. 7. The diagram shows the Source composite curve
1,400,000 Sink composite curve
result of a preference for condensed water
from Evaporators 1 and 2, as an additional 1,200,000
practical constraint. This option is pre-
1,000,000
ferred due to their relatively lower contain-
Load, gr/hr

ment concentrations. 800,000


The resulting network scheme is then
connected to the internal Excel-based 600,000
plant mass balance model, producing a
wastewater stream that is sent to the evap- 400,000
orator model. Although the allowed con-
200,000
centration of the sinks is obtained from
the plant as the maximum concentration 0
allowed, this step confirms the effect of the 0 1,000 2,000 3,000 4,000 5,000 6,000
Flow, tph
recycle streams to the product quality. The Freshwater intake
mass balance model shows that the pro- FIG. 6. Water pinch analysis using source-sink composite curve method.
posed recycled water streams do not have
a significant impact on product quality. Optimum water network with Evaporator 1 + Evaporator 2
The produced wastewater stream now 650 kg/hr
has a slightly different composition than 520 Unit 1
910 kg/hr
before, so it is put into the proprietary Freshwater 390
650 kg/hr
model of the multiple-effect evaporator. Water cost (thousand €/yr): 7 Unit 2
130
This is also done to see if the concentra- 2,000 kg/hr 260
Evaporator 1
tion of contaminants in the condensed wa- 50 50 kg/hr 603 ppm
ter streams from the evaporators is affect- Unit 3 1,210 kg/hr to WWTP
ed. The contaminant concentration of the 292 605
663 Wastewater cost
condensed water is not affected, making 1,300 kg/hr (thousand €/yr): 10
637 Unit 4 Total cost
1,900 kg/hr 160
the proposed water network acceptable. Evaporator 2 (thousand €/yr): 17
Reduction: 78%
595 1,200 kg/hr
Wash water to centrifuge. In existing 508 Unit 5
practice, the centrifuge unit is washed with 1,000 kg/hr 1,000
Evaporator 3
freshwater due to salt accumulation. This 800 kg/hr
water is then sent to the second evapora- Unit 6
tor. In a pinch analysis point of view, this 50 kg/hr 50
situation leads to three consequences, all Dryer
of which are related to one another:
• Higher freshwater consumption FIG. 7. Optimum water network with reusing only condensed water from Evaporators 1 and 2.

Hydrocarbon Processing | DECEMBER 2015 77


Water Management

Centrifuge with wastewater


Collection Collection
To WWTP vessel 1 vessel 2
Unit 1

Unit 2 Hot oil out Hot oil in

Plant
Fresh water
To WWTP
1st 2nd 3rd
Unit 3 effect effect effect

CW out CW in
Unit 4

Unit 5 Pit
Centrifuge Salt
(Unit 6)
Concentrated organics

FIG. 8. Option of not using freshwater to clean the centrifuge.

1,600,000 Takeaway. A water pinch analysis study


Wastewater discharge
Source composite curve has been performed to minimize fresh-
1,400,000 Sink composite curve water consumption and wastewater dis-
1,200,000 charge by a specialty chemical plant. This
method has proven that the systematic
1,000,000 approach of pinch analysis can give bet-
ter insight into the plant improvement.
Load, gr/hr

800,000 In this particular case, it shows that the


600,000
targets of freshwater consumption and
wastewater discharge are lower than ini-
400,000 tially proposed within the plant. It also
shows, without going into too much de-
200,000 tail, that using wastewater directly from
0
the pit to wash the centrifuge does not
0 1,000 2,000 3,000 4,000 5,000 6,000 reduce freshwater consumption.
Flow, tph The results of the study and the math-
Freshwater intake
ematical optimization of the developed
FIG. 9. Water pinch analysis without the centrifuge. superstructure provide a water network
solution that can be applied in practice.
TABLE 2. Sink streams The solution shows that practical con-
straints are considered—e.g., the use of
Number Flow, kg/hr Concentration, ppm Load, mg/hr Stream name
condensed water from Evaporators 1 and
1 650 10 6,500 Unit 1
2 only.
2 650 20 13,000 Unit 2
LITERATURE CITED
3 50 50 2,500 Unit 3 1
Smith, R., Chemical Process: Design and Integration,
4 1,300 300 390,000 Unit 4 John Wiley & Sons, Hoboken, New Jersey, 2005.
5 1,200 300 360,000 Unit 5
2
El Halwagi, M., Process Integration, Academic Press,
Waltham, Massachusetts, 2006.
6 800 500 400,000 Unit 6 3
Amminudin, K. A., A. Buang, H. Hawari, P. N. F.
M. Khamaruddin, Z. A. Putra and M. N. Sofian,
tively cleaner; or, in other words, it has a sumption and wastewater discharge do “Consider water reuse options for your facility,”
very dilute salt concentration, making it not change; the reason is that the centri- Hydrocarbon Processing, December 2008.
useful to dissolve this salt accumulation. fuge unit can accommodate a higher con- ZULFAN ADI PUTRA is a process
The resulting water stream and the main taminant concentration. So, instead of us- engineering and process technology
wastewater stream input are returned to ing freshwater, any other condensed water consultant at TMC Chemical in The
Netherlands. He has been involved
the first evaporator (FIG. 8). can be used to wash the centrifuge prior in a wide range of work activities,
This option appears to bring an ad- to discharge. A holistic and systematic ap- such as conceptual process designs,
ditional benefit. Based on this change, proach involving water pinch analysis, via feasibility studies, basic engineering
source-sink composite curves are drawn source-sink diagram, can tackle this issue design and plant performance evaluations, which
include process simulation, optimization and
to ascertain if this option is useful. FIG. 9 quickly without going into the details of debottlenecking studies. He holds a PDEng degree
shows that the targets for freshwater con- the water network. from Technische Universiteit of Eindhoven.

78 DECEMBER 2015 | HydrocarbonProcessing.com

You might also like