(Microsoft PowerPoint - Robin Smith Presentation 2 [Kompatibilitetsl ...
Post on 12-Feb-2017
222 Views
Preview:
Transcript
Centre for Process Integration © 2012
From Energy to Capital and Economic Optimisation
Robin Smith
University of Manchester
• Objective
• Develop a methodology for
optimising the design or operation
of heat-integrated processes in
order to:
- Maximise profit
- Maximise yield of valuable products
- Maintain product quality
1. Energy and Capital Cost Targets
Capital Energy Trade-Offs
The correct setting for ∆Tmin is economic.
T
H
2
∆Tmin2
T
1
H
∆Tmin1
COST
Total
Capital
OPT
1 2
Energy
∆Tmin
We can set energy cost targets
as a function of ∆Tmin.
Energy
Cost
∆Tmin
But what about capital cost ?
CAPITAL COST
Area
No Units
No ShellsMaterials of
Construction
Equipment
Type
Pressure
Rating
intervals
Network Area
Amin = ∑ ∑
We can set overall area targets
1
∆TLM
q
hstreams
T
H
7
6
5
4
3
2
1
.
Number of Units
Nabove = Sabove - 1
PINCH
Nbelow = Sbelow - 1
NMER = (Sbelow - 1) + (Sbelow -1)
S = number of streams including utilities
We can also account for
• No of shells
• Materials of Construction
• Equipment Type
• Pressure Rating
Trading Off Energy and Capital Cost Targets
∆Tmin ∆Tmin
∆Tmin∆Topt
Capital
Cost
Energy
Cost
Cost
Trading off Energy and Capital Cost Targets
• Only applies to new design
• Capital cost unreliable, even in new design� Network structure is necessary for more
reliable cost estimates
BETTER APPROACH FOR CAPITAL COST
• Automated design for new networks
BUT
2. Heat Exchanger Network Retrofit
BUT, what about Retrofit of Heat Recovery Networks?
How can we tackle the retrofit problem more
effectively?
E14 E10E11E13 E12
DESALTER
FURNACE
FLASH
E7 E8 E9
E6
E2 E1E3
E5 E4
CIT
Vac.Res.
Atm.MPA
Atm.2SS
Atrm.TPA
Vac.PA
Vac.PA
Atm.3SS
CrudeOil
E14 E10E11E13 E12
DESALTER
FURNACE
FLASH
E7 E8 E9
E6
E2 E1E3
E5 E4
CIT
Vac.Res.
Atm.MPA
Atm.2SS
Atrm.TPA
Vac.PA
Vac.PA
Atm.3SS
CrudeOil
INFEASIBLE REGION
2 mods
3 mods
Retrofit Approach
ExistingHENAexist
Eexist
0 mods
1 mod
Search for structural changes with best cost-effective designs
A
E
Stochastic Optimization can be used to Optimize Structural Options
Simulated annealing moves
Repipe HX
ResequenceHX
Add newHX
Add/changestream split
AND the continuous variables
• Gives the designer greater control over the complexity of the retrofit design
• Allows a range of retrofit options to be identified
• Because of the increased control allows a more practical approach to retrofit
This Approach to Retrofit
BUT....
• Obtaining cost effective retrofits is still our biggest challenge
• Need to minimize piping modifications and civil engineering
Retrofit of a crude oil distillation preheat train
Case Study
Existing Heat Exchanger Network
Hx no. Area(m2)
1 3262 333
3 10
4 5
5 167
6 150
7 305
8 317
9 16
10 38
11 20
12 15
13 154
14hu 118
15 39
16hu 26
17hu 14
18 23
19hu 5
21hu 722 164
23cu 45
24cu 1,151
25cu 55
26cu 119
27cu 234
28cu 82
Total 3,939
1210864
119753
2
1
M1
1314
13
13
11
11
12
12
M2
15182221
1
1
2
2
7
7
8
8
M4
5
5
6
6
M5
15
15
19
3
3
18
18
25
26
16
4
4
22
22
27
24
23
17
9
9
10
10
M3
28
Optimized Heat Exchanger Network
4
4
4 121086
11753
2
1
1314
13
13
11
11
12
12
M2
15182221
1
1
2
2
7
7
8
8
M4
5
5
6
6
M5
15
15
19
3
3
18
18
25
26
16
22
22
27
24
23
17
9
9
10
10
M3
28
4
4
Additional area
NEW HX
Repipe
4
4
Summary of Case Study
Hot Utility (MW)
Existing Optimised
99.5 96.4
Saving
3.1
Total operating Cost(106 £/yr)
7.46 6.101.36
(18.2 %)
Cold Utility (MW) 83.8 78.8 5.0
Additional Area (m2) - 1022 -
� Problem, too much additional area requiredthroughout the network
• Retrofit design often involves too many modifications to the network equipment and structure.
• Practical implementation of the additional area in an existing network can be difficult due to topology, safety and downtime constraints.
Problems in Retrofit Design
• Heat exchanger networks retrofit usually requires increasing heat exchanger area in various places in the network.
• Can be expensive to add additional area (especially pipework and civil engineering).
Retrofit of HENs
• Need to minimize piping modifications and civil engineering, exploit new heat transfer technology, etc.
Overcoming Retrofit Problems
New Technologies
Tube Inserts(©HiTran)
• Heat transfer enhancement is one way of increasing the heat transfer in heat exchange equipment.
• General reluctance to adopt HTE due to fear of increased fouling (in practice, it generally decreases fouling)
• It can give heat exchangers a higher heat transfer coefficient, and make exchangers smaller, therefore cheaper and can give greater energy efficiency for processes.
Heat Transfer Enhancement (HTE)
Crude Oil Fouling Research [University of Bath]
Foulin
g r
esis
tance [m
2K
/ W
* 1
04]
Time [hr]
Tube with matrix Element velocity 0.5m/sTwall = 218 °C
Time [hr]
Tube without matrix Elementvelocity 0.5m/sTwall = 216 °C
Arabian light crude oil
Advantageous effects:• Lower tube-wall temperature for same duty• Reduction in fluid volume which is heated above bulk temperature• Reduction in wall fluid residence time• Suppression of nucleation at the surface• Increased shear rate at the wall (higher removal rate)
Benefit of Applying HTE in HENsRetrofit
• Implementation of HTE is relatively simple.
• This means reduced civil and pipe work costs.
• The duration of the modifications also decreases, which means the modifications may be implemented during a normal shut down period to avoid production losses.
• Implementation of HTE is generally much cheaper than additional area.
Further Improvements to Performance
How can we increase the performance further?
Simultaneous process and heat recovery optimization
3. Towards Profit Optimisation
Heat-integrated distillation system optimisation
Develop a methodology for optimising the operating conditions of a crude oil distillation system
ANN model to simulate distillation unit
Model to simulate HEN
Heat-integrated distillation system optimisation
Distillation system with current operating conditions and fixed
structure
Distillation system with optimised operating conditions
SIMULATED ANNEALING ALGORITHM
Distillation model
HEN model
Economic model
Distillation system simulation
Penalty functions
Objective function
Product quality specsSystem feasibility
• SA parameters
• Upper and lower bounds of optimisation variablesDistillation system
parameters
Case study
Existing column and HEN- Light crude (Bombay crude)
- Allow product flow rates to change
- Keep product specifications
Case study - Crude oil distillation model
ANN model results – Validation (Bombay crude)
Profit Optimization
Objective function
Net profit = Product value – cost of crude oil feed
– annualized capital cost – operating costs
Constraints• Hydraulic limits of distillation columns• maximum pump-around flow rate increase• Product Specifications• Product boiling curve properties of both atmospheric
column and vacuum column• Product flow rate of vacuum column• Steam flow rates
Constraints
Distillation
•Product T5% and T95% TBP points variation of less than ±10°C from the base case values.
•Calculated column diameters less or equal than existing diameters
HEN
•Minimum temperature approach ≥ 25°C
•Stream energy balance constraint
•No topology modifications
•Calculated area less or equal than installed area
Operational optimisation results
Item Base caseMinimisation of
energy consumption
Maximisation of net profit
Hot utility (MW) 46.6 44.4 44.2
Cold utility (MW) 74.7 73.6 72.5
Hot utility costs (MM$/y) 7.0 6.6 6.6
Cold utility costs (MM$/y) 0.4 0.4 0.4
Utility costs (MM$/y) 7.4 7.0 7.0
Operational optimisation results
Item Base caseMinimisation of
energy consumption
Maximisation of net profit
Product revenue (MM$/y) 2 879.2 2 879.2 2 894.2
Crude oil costs (MM$/y) 2 866.9 2 866.9 2 866.9
Steam cost (MM$/y) 1.7 1.7 1.7
Utility costs (MM$/y) 7.4 7.0 7.0
Total profit (MM$/y) 3.2 3.6 18.5
Summary
• Given a combination of process and HEN model, profit optimisation possible
�Operational optimisation (detailed HEN model)
�Retrofit (detailed HEN model)
�New design (automated design or HEN target)
• Especially effective in retrofit if products arebottlenecked
• Such profit optimization has been appliedindustrially
Acknowledgments
• Lluvia M. Ochoa-Estopier, Megan Jobson (University of Manchester)
• Martin Gough
(CalGavin Ltd)
• Financial support is gratefully acknowledged from the EC FP7 project “Efficient Energy Integrated Solutions for Manufacturing Industries – EFENIS”, Grant Agreement No 282789.
Thank you!
top related