0 Up votes0 Down votes

1 views15 pagesOptimization

Dec 10, 2018

© © All Rights Reserved

PDF, TXT or read online from Scribd

Optimization

© All Rights Reserved

1 views

Optimization

© All Rights Reserved

- em8779-e[1]
- BESO3D
- SRS-298
- 20.IJASRAPR201720
- Paper 016
- dataleakagedetection123-131216094254-phpapp01
- MATH5.MODULE-BODY (1).docx
- Scheduling-Optimisation-of-Chemical-Process-Plant.pdf
- Linear Programming_problems_EN_29-5-2012.pdf
- Linear Programming Exercises
- Optimization Handout
- Linear-programming.pptx
- 2-78-1424077448-3. management - OPTIMIZING CEMENT DISTRIBUTION IN THE NIGERIAN CEMENT MANUFACTURING- Dr Nwekpa K C (1).pdf
- EE16
- WCSMO4 FAIPA SAND Herskovits Mappa Juillen 2001
- Lot-sizing in flow-shop with energy consideration for sustainable manufacturing systems
- Us 8082167
- Fuzzy Linear Programs With Trapezoidal Fuzzy Numbers
- Solving a Gas-lift Optimization Problem by Dynamic Programming
- 8 Chapter 3 Graphical Method

You are on page 1of 15

OPTIMUM DESIGN and DESIGN STRATEGY

• Optimization is the process of improving an existing situation, device, or system such as a chemical process.

o Set up an optimization problem

o Quantify the value of a potential improvement

o Identify quickly the potential for improvement

o Identify the constraints, barriers, and bottlenecks for improvement

o Choose an appropriate procedure to find the best change

o Evaluate the result of the optimization

• Terms and Definition:

o Decision variables (also called design variables) – are those independent variables over which the engineer has some

control. These can be continuous variables such as temperature or discrete (integer) variables such as number of

stages in a column.

o Objective function – is a mathematical function that, for the best values of the decision variables, reaches a minimum

(or maximum). Thus, it is the measure of value or goodness for the optimization problem.

§ If it is a profit, one searches for its maximum.

§ If it is a cost, one searches for its minimum.

§ There could be more than one objective function for a given optimization.

o Constraints – are limitations on the values of decision variables.

§ These may be linear or non-linear, may involve more than one decision variable

§ It is called an equality constraint if it is written as an equality involving two or more decision variables

§ Example: a reaction may require a specific oxygen concentration in the combined feed to the reactor. The

mole balance on the oxygen in the reactor is an equality constraint.

§ It is called an inequality constraint if it is written as an inequality involving one or more decision variables

o

§ Example: the catalyst may operate effectively below 400 C, or below 20 Mpa.

§ An equality constraint effectively reduces the dimensionality (the number of truly independent decision

variables) of the optimization problem.

§ Inequality constraints reduce (& often bound) the search space of the decision variables.

o Global optimum – is a point at w/c the objective function is the best for all allowable values of the decision variables.

There is no better acceptable solution.

o Local optimum – is a point from w/c no small, allowable change in decision variables in any direction will improve the

objective function.

o Linear programming – the optimization method that is used if the objective function is linear in all decision variables

and all constraints are linear.

o Non-linear programming – optimization method that is used in other optimization problems

o Quadratic programming – a nonlinear optimization used if the objective function is second order in the decision

variables and the constraints are linear

o Mixed integer – used for optimization problems involving both discrete and continuous decision variables

Optimization is usually a dynamic, creative activity involving brainstorming, exploring alternatives, and asking “what if?”

• Methods:

1. Classical Optimization 3. Dynamic Programming

a) One variable – differential calculus 4. Method of Lagrange

b) Two or more variable – partial differentiation Multipliers

2. Linear Programming 5. Method of Steepest

§ Graphical Ascent or Descent

§ Simplex method 6. Geometric Programming

• General Procedure for Determining Optimum Conditions

a. Determine what factor is to be optimized (objective function).

§ Typical factors are: total cost per unit of production or per unit time

• Profit

• Amount of final product per unit time

• Percent conversion

1

ChE 522: Chemical Engineering Plant Design and Economics Lecture Notes

b. Develop relationships showing how the different variables (decision variables & constraints) involved affect

the chosen factor

c. Combine relationships graphically or analytically to give the desired optimum conditions

• Procedure with One Variable

o Graphically, the optimum condition is the maximum point or the minimum point on the curve obtained by plotting the

objective function vs decision variable.

Optimum Design for One Variable

Optimum diameter

st

o Analytically, take the 1 derivative of the objective function, equate to zero and solve for the unknown variable.

The value of the variable occurs at an optimum point or point of inflection.

nd

o Take the 2 derivative and check if it is

§ > 0, the value occurs at a minimum

§ < 0, the value occurs at a maximum

§ equal to zero, the value occurs at the point of inflection

• Procedure with Two or More Variables

A. Graphical

§ Two dimensional (see fig.)

The factor being optimized is plotted against

one of the independent variable (x) w/ a

second variable (y) held at a constant value.

Plot the points, the optimum value of x and y

occurs at the minimum point on the curve

§ Three dimensional

Solve first for optimum value of CT, x, and y at

one constant value of z then plot z versus the

optimum values of CT.

B. Analytical Method

§ Take partial derivatives, equate to zero.

§ Solve the two equations simultaneously.

§ If more than two independent variables

are involved, same procedures must be

employed.

Graphical determination of optimum conditions w/ 2 or more independent variables

Optimum Production rates in Plant Operation

• One of the most important variables in any plant operation is the amount of product produced per unit time.

• Production rate depends on many factors such as:

o the number of hours in operation per day, per week, or per month

o the load placed on the equipment

o the sales market available

2

ChE 522: Chemical Engineering Plant Design and Economics Lecture Notes

• Analysis of the costs involved under different conditions & considerations of these & other factors allows determination of

the optimum rate of production also called economic lot size.

• Design factors/considerations and actual plant operation may differ from each other and that this may affect production

rates.

• Price and market conditions do not remain constant so that even if the design engineers have based their recommendation

on a similar item, the actual data may not always coincide with the design data.

• Classification of total product cost

o Operating costs – costs depending on the rate of production & include expenses for direct labor, raw materials,

heat, power, supplies, & other similar items w/c are function of the amount of material produced.

§ Basis of operating cost is per unit of production

Types of operating costs

§ Minimum expenses for RM, labor, power, etc.

§ Extra expenses (also known as super-production costs) due to increasing the rate of production.

§ Example of super-production costs – extra expenses caused by overload on power facilities, additional

labor requirements, decreased efficiency of conversion

Organization costs – expenses due for directive personnel, physical equipment, and other services or facilities w/c must be

maintained irrespective of the amount of material produced.

n

Superproduction cost per unit of production = mP eqn. 01

Where: P – rate of production (total units produced/time)

m, n – constants

Let: h be the operating costs, constant per unit of production

Oc – organization costs per unit time

cT – total product cost per unit of production

Then:

n

cT = h + mP + Oc ; $ _ eqn. 02

P unit

From equations 1 & 2, the following can be derived:

n

CT = cTP = h + mP + Oc P where CT - total product cost; $/time

P r – profit per unit of production

n

r = s – cT = s - h + mP + Oc ; $ s – selling price/unit production

P unit R - profit per unit time

n

R = rP = s - h - mP - Oc P ; $ eqn. 03

P time

Optimum Production Rate For Minimum Cost Per Unit of Production

• The determination of the production rate that will give the least cost is very important.

• Graphically, plotting the total product cost per unit against production rate will show the minimum product cost and the

production corresponding to it is the optimum production rate.

• Analytically, by taking the first derivative of eqn. 2 and equating to zero, an optimum production rate can be obtained. i.e;

n-1 1/(n+1)

dcT = nmPo – Oc = 0 ; Po = Oc where Po – optimum production rate

2

dP Po nm

Optimum Production Rate For Maximum Total Profit

• In some cases, the amount of money earned over a given time period is more important than the amount of money earned

for each unit of product sold.

• The production rate that will give the maximum profit per unit of time may differ from the production rate that will give the

minimum cost per unit of production.

• Graphically, a plot of profit per unit time against production rate goes to a maximum.

• Analytically, by taking the first derivative of eqn. 03 and equating to zero, an optimum production rate is obtained, i.e.

n 1/n

dR = (s-h)Po – m(n+1)Po = 0 ; Po = s – h

dP m(n +1)

3

ChE 522: Chemical Engineering Plant Design and Economics Lecture Notes

Optimum Conditions in Cyclic Operations

• Some operations are cyclic w/c involve shutdown for some activities.

• Such type of processes occurs in batch processes or when rate of production decreases with time.

• In true batch processes, no product is obtained until the unit is shutdown for discharging.

• A good example for batch process is the operation of filter presses such as the PFFP and LF.

• In semi-continuous cyclic operations, the product is delivered continuously but the rate of delivery decreases w/ time.

• The variable total time required per cycle must therefore be considered in determining optimum conditions.

• For such operations, relationships similar to the following can be developed to express overall factors. Basis: 1 cycle

Total annual cost = (operating & shutdown costs/cycle)(cycles/year) + AFC

Annual production = (production/cycle)(cycles/yr)

Cycles per year = operating + downtime/yr .

operating + downtime/cycle

Problem: Determination of profits at optimum production rates (ex.2/352)

1.2

A plant produces pumps at the rate of P units per day. The variable costs per pump have been found to be $47.73 + 0.1P . The

total daily fixed charges are $1750, and all other expenses are constant at $7325/day. If the selling price per pump is $173,

determine (a) the daily profit at a production schedule giving the minimum cost per pump. (b) the daily profit at a production

schedule giving the maximum daily profit. (c) the production schedule at the break- even point.

1.2

Given: Variable cost; $/pump = 47.73 + 0.1P

FC = $1750 + $7325 = $9075/d

Selling price; $/pump = 173

Req’d: a) daily profit at a production giving minimum cost per pump

b) daily profit at a production giving maximum daily profit

c) production at break even point

Soln: P = units/d

1.2

a) Objective function is cT: cT = VC + FC = 47.73 + 0.1P + 9075 ; $ .

P unit

0.2 2.2

Minimize cost/unit: dcT = 0.1(1.2)Po - 9075 = 0 ; 0.12Po = 9075 Po = 165.0 units/d

2

dP Po

Daily profit (at optimum production giving minimum cost) is determined from

1.2

R = (s – cT)Po = {173 – [47.73 + 0.10(165) + 9075/165]}165

R = $4035.54/d

b) objective function is R:

1.2

R = (s – cT)P = [173 – (47.73 + 0.1P + 9075/P)]P

1.2

Maximize profit/d: dR = 173 – 47.73 – 0.1(2.2)Po = 0

dP

1.2

125.27 = 0.22Po Po = 197.78 = 198 units/d

Daily profit at Po is:

1.2

R =[173 – (47.73 + 0.1(198) + 9075/198)]198 = $4439.25/d

c) @ break even point, profit = 0:

1.2

[173 – (47.73 + 0.1Po + 9075/Po)]Po = 0 Solving; Po = 88 units/day

4

ChE 522: Chemical Engineering Plant Design and Economics Lecture Notes

Determination of conditions for minimum total cost in a batch operation:

An organic chemical is being produced by a batch operation in which no product is obtained until the batch is finished. Each cycle

consists of the operating time necessary to complete the reaction plus a total time of 1.4 hr for discharging and charging. The

0.25

operating time per cycle is equal to 1.5Pb hr, where Pb is the kg of product/batch. The operating costs during the operating period

are $20/hr and the costs during the discharge-charge periods are $15/hr. The annual fixed costs for the equipment vary with the

0.8

size of the batch as follows: CF = 340Pb $/yr. Inventory & storage charges may be neglected. If necessary, the plant can be

operated 24 hr/d for 300 days/yr. The annual production is 1M kg of product. At this capacity, raw material and miscellaneous

costs, other than those already mentioned, amount to $260,000/yr. Determine the cycle time for conditions of minimum total cost

per year.

0.25

Given: tdischarge&charge = 1.4 hr ; operating time = 1.5Pb hr

Optg costs = $20/hr, shutdown cost = $15/hr,

0.8

AFC = CF = 340Pb $/yr, other FC = $260,000/yr

Production = 1 M kg

Req’d: cycle time for conditions of minimum total cost per year.

Sol’n: Pb = kg production/cycle ; Let CT = total annual costs

CT = (optg costs/cycle)(cycles/yr) + (shutdown costs/cycle)(cycles/yr) + Total Annual Fixed Costs

6 6

Cycles/yr = annual production = 1x10 kg prod’n/yr = 1x10 cycles

production/cycle Pb kg prod’n/cycle Pb yr

0.25 0.25

Operating costs/cycle = ($20/hr)(1.5Pb )hr/cycle = 30Pb $/cycle

Shutdown costs/cycle = ($15/hr)(1.5 hr/cycle) = $21/cycle

0.25

Cycle time = operating time + downtime = 1.5 Pb + 1.4 hrs

0.25 6 6

CT = 30Pb $ 1x10 cycles + $ 21 1x10 cycles

0.8

Cycle Pb yr cycle Pb yr + 340Pb + 260000

6 6 0.8 st

CT = 30x10 + 21x10 + 340Pb + 260,000 Minimize TAC to give optimum Pb at min cost, 1 derivative = 0

0.75

Pb Pb

6 6

d(CT) = -(0.75)30x10 - 21x10 + 0.8(340) = 0 ;

1.75 2 0.2

dP Pb Pb Pb

6 -1.75 6 -2 -0.2

- 22.5x10 Pb – 21x10 Pb + 272 Pb = 0 Pb,o = 1626 kg/cycle

0.25

@ Pb,o = 1626 kg/cycle; cycle time = 1.5 Pb + 1.4

0.25

cycle time = 1.5(1626) + 1.4 = 10.925 hrs/cycle (Ans)

6

total time used = 10.925 hrs 1x10 cycles = 6718.9 hrs/yr

yr cycle 1626 yr

total available time/yr = 300(24) = 7200 hrs/yr

Cycle time for maximum amount of production from a PFFP:

Tests w/ a PFFP, operated at constant pressure, have shown that the relation between the volume of the filtrate delivered and the

2 3

time in operation can be represented as follows. Pf = 18(θf + 0.11) where Pf is the m of filtrate delivered during θf hrs of filtering

5

ChE 522: Chemical Engineering Plant Design and Economics Lecture Notes

th

time. The cake formed in each cycle must be washed w/ an amount of water equal to 1/16 times the volume of the filtrate

delivered per cycle. The washing rate remains constant and is equal ¼ of the filtrate delivery rate at the end of the filtration. The

time required per cycle for dismantling, dumping and reassembling is 6 hrs. Under the conditions where the preceding information

applies, determine the total cycle time necessary to permit the maximum output of filtrate during each 24 hrs.

Given:

Pf = volume of filtrate delivered/cycle Vw = Pf ; rw = ¼ rf & rf = 1 dPf

1/2 3

Pf = 4.24(θf+0.11) m /cycle 16 4 dθf

θf = filtration time/cycle θw = washing time/cycle

θd,d,r = 6 hrs/cycle

Reqd: optimum cycle time for maximum output of filtrate/day

Soln: objective function is to maximize volume of filtrate per day, Vf

3

Vol delivered, Vf = N, cycles x Pf, m . ; Cycles = 24 hrs/day .

day day cycle day cycle time; hrs/cycle

Let total cycle time be θT : θT = θf + θd,d,r + θw

1/2

θw = Vw ; rw = ¼ rf and rf = dPf = d (4.24)( θf+0.11)

rw dθf dθf

-1/2

rf = ½ (4.24)( θf+0.11) (1)

3

rf = 2.12 ; rw = 1 2.12 = 0.53 m

1/2 1/2 1/2

(θf+0.11) 4 (θf+0.11) (θf+0.11) hr

1/2 1/2 1/2

θw = Vw = Pf(θf+0.11) = 4.24(θf+0.11) (θf+0.11) = (θf+0.11)

rw 16(0.53) 16(0.53) 2

θT = θf + θddr + θw = θf + 6 + θf + 0.11 = 1.5 θf + 6.055 hrs .

2 cycle

cycles = 24 hrs/d .

day (1.5 θf + 6.055) hrs/cycle

1/2

volume delivered = 24 4.24(θf+0.11)

day 1.5 θf + 6.055

1/2 3 1/2 -1

Vf = 101.76(θf+0.11) ft = [101.76(θf+0.11) ](1.5θf + 6.055)

(1.5θf + 6.055) day

-1/2 -1 1/2 -2

dVf = 101.76 (½)(θf,o + 0.11) (1)(1.5θf,o + 6.055) + 101.76(θf,o+0.11) (-1)(1.5θf,o+6.055) (1.5) = 0

dθf

1/2

50.88 = 152.64(θf,o+0.11)

1/2 2

(θf + 0.11) (1.5θf+6.055) (1.5θf,o+6.055)

50.88(1.5θf,o+6.055) = 152.64(θf,o + 0.11) ; θf,o = 3.82 hrs

Cycle time, θT = 1.5θf + 6.055 = 1.5(3.82) + 6.055 = 11.785 hrs/cycle

Semicontinuous Cyclic Operations

• Product is delivered continuously but the rate of delivery decreases with time owing to:

o scaling,

6

ChE 522: Chemical Engineering Plant Design and Economics Lecture Notes

o collection of side product,

o reduction in conversion efficiency

o or similar other causes.

• It is necessary to shutdown operation periodically in order to restore the original conditions for high production rates.

• The optimum cycle time can be determined for conditions such as maximum amount of production per unit time or minimum

cost per unit of production.

Scale Formation:

• Solids deposit on the heat-transfer surfaces (scales are formed)

• Continuous formation of scales

o causes a gradual increase in the resistance to flow of heat

o and thereby reducing the rate of heat transfer

o and rate of evaporation if the same temperature difference driving forces are maintained.

• These condition requires shutdown for cleaning

• The time for cleaning is included in the optimum operation time

• Scaling occur especially when the feed contains a dissolved material that has inverted solubility

• Inverted solubility means the solubility decreases as the temperature of the solution is increased.

• Fro such a material, the solubility is least near the heat-transfer surface where the temperature is greatest.

• Thus, the solids crystallize out from the solution forming a scale and are deposited on the heat-transfer surface.

• Examples of common materials causing scaling are: calcium sulfate, Ca(OH)2, Na2CO3, Na2SO4 and other calcium salts.

Magnesium salts also form scales.

• When true scale formation occurs, equations similar to the following applies:

1 = aθb + d where: a & d – are constants; θb – operating time/cycle

2

U U – overall heat transfer coefficient

• Other appropriate equations can be used when it is not convenient to determine the overall heat transfer coefficient.

• The rate of heat transfer at any time is: separate variables & integrate:

Q θb

dQ = UAΔt = AΔt . dQ = A Δt 1 dθb

1/2 0 0 1/2

dθb (aθb + d) (aθb + d)

1/2 1/2

Q = 2AΔt [(aθb + d) – d ] where Q – heat transferred/cycle

a Δt – temperature gradient; A – heat transfer area

Cycle Time for Maximum Amount of Heat Transfer

• Equations that relate heat transfer and time can be used for finding the cycle time that will permit the maximum amount of

heat transfer.

• Each cycle consists of an operating or boiling time, θb and time for empting, cleaning & recharging is θc

• Total time per cycle: hrs/cycle = θt = θb + θc

• Total time used for actual operation, emptying, cleaning & refilling = H (hrs)

• The number of cycles during the overall operation is:

Cycles = H hrs

(θb + θc ) hrs/cycle

• The total amount of heat transferred during the operation is:

QH = (Q/cycle) x (cycles/H,hr) = amount of heat/hr

1/2 1/2

QH = 2AΔt (aθb + d) – d H .

a (θb+θc)

• Graphically, a plot of total amount of heat transferred versus θb will show a maximum at the optimum value of θb.

st

• Analytically, the 1 derivative of QH w/ respect to θb is set to zero and θb is solved.

1/2

dQh = 0 ; θb = θc + 2 (adθc)

dθb a

• Use of tangential method for finding optimum conditions

o Used for constant cleaning time (θc).

o A plot of the amount of heat transferred vs boiling time is prepared.

7

ChE 522: Chemical Engineering Plant Design and Economics Lecture Notes

o A tangent to the curve starting from θc (cleaning time) w/ slope equal to Q/(θb + θc) define the optimum point.

o The point of tangency will give the optimum value of the boiling time per cycle for conditions of maximum amount

of heat transferred.

See figure 11-5, page 358 for an example

o

Cycle Time for Minimum Cost per Unit of Heat Transfer

• Assumption are as follows:

o An evaporation unit of fixed capacity is available

o A definite amount of feed and evaporation must be handled each day

o The total cost for one cleaning & inventory charge is constant no matter how much boiling time is used.

• Determine the cycle time that will permit operation at the least cost.

• The total cost includes

o Fixed charges on the equipment & fixed overhead expenses

o Steam, materials & storage costs, proportional to the amount of feed & evaporation

o Expenses for direct labor during the actual evaporation operation

o Cost of cleaning

• First two costs are independent of time

• An optimum cycle time can be found by minimizing the total costs for cleaning & for direct labor during the evaporation

• Let Cc be the cost for one cleaning, Sb be the direct labor cost per hour during operation

• The total variable costs during H hours of operating and cleaning time is:

CT, for H hrs = (Cc + Sbθb) H .

θb + θc

• Substituting the term H/(θb+θc) from QH equation:

CT, for H hrs = aQH(Cc + Sbθb) .

1/2 1/2

24AΔt[(aθb + d) – d ]

• CT, for H hrs can be minimized to obtain optimum cycle time that will give the least cost

• Graphically, CT can be plotted vs θb, minimum point determines optimum cycle time

st

• Analytically, 1 derivative can be taken and equated to zero and optimum θb can be solved.

1/2

dCT = 0 ; θb,o = Cc + 2 (adCcSb) ; time/cycle

dθb Sb aSb

• Checking total time θt:

1/2 1/2

θt = θb + θc = 2AH’Δt[(aθb,opt + d) – d ]

aQH

where H’ – total time available for operation, emptying, cleaning & recharging

Accuracy and Sensitivity of Results:

• The principles illustrated in the preceding discussions shows general concepts on the determination of optimum conditions.

• Other factors such as cost due to taxes, time value of money, capital, efficiency or inefficiency of operation & special

maintenance were not emphasized.

• Such factors may influence the optimum conditions and must therefore be taken into account in the final analysis.

• The engineer must practically understand and recognize when to take into account such factors and when the added

accuracy obtained by accounting them is not worth the difficulty they cause in the analysis.

Fluid Dynamics (optimum Economic Pipe Diameter)

• Investment for piping and pipe fittings is an important part of the total investment for a chemical plant.

• It is therefore necessary to choose the size of pipes that will give the minimum cost for pumping & fixed charges.

• The use of increased pipe diameter will cause an increase in the fixed charges for the piping system & decrease in the

pumping or blowing charges.

• An optimum pipe diameter must exist.

• This optimum condition can be found at the point at w/c the sum of pumping & blowing costs plus fixed charges on the

piping system is a minimum.

• Pumping and Blowing Costs:

For incompressible fluid through a pipe of constant diameter:

2

o W = 2fV L(1+J) + B where: W- mechanical work added (ft-lbf/lbm)

gcD

8

ChE 522: Chemical Engineering Plant Design and Economics Lecture Notes

For turbulent flow (Nre > 2100) & for new steel pipes: f – fanning friction V – ave fluid linear velocity,ft/s

o f = 0.04 : NRe = DVρ/µ L – length of pipe, ft

0.16

(NRe) J – friction loss due to fittings expressed as eq loss in straight pipes

gc – constant

For laminar flow (NRe < 2100) f = 16/NRe B – constant taking all other factors of the ME balance

Annual pumping cost for turbulent flow is given by:

2.84 0.84 0.16

o Cpumping = 0.273qf ρ µ K(1+J)Hy + B’

4.84

Di E

Annual pumping cost for viscous flow is given by: Di – inside pipe dia, inch

2

o C’pumping = 0.024qf µc K(1+J)Hy + B’

4

Di E

µc – fluid viscosity, cP ; K – constant of electrical energy, $/kWh

Hy – hours of operation per year; E – efficiency of motor & pump (fraction)

B’ - a constant independent of Di

Cpumping – pumping cost ($/yr) per foot of pipe length when flow is turbulent

C’pumping – pumping cost ($/yr) per foot of pipe length when flow is viscous

• Fixed Charges for piping System

n

o The purchased cost is cpipe = XDi

n

o The annual cost for installed piping system is : Cpipe = (1+F)XDi KF

o Where: cpipe – purchase cost of new pipe/ft pipe length, $/ft

o X – purchase cost of new pipe/ft of pipe length if pipe dia is 1 in, $/ft

o N – a constant w/ value dependent on type of pipe

o Cpipe – cost for installed piping system, $/yr/ft of pipe length

o F – ratio of total costs fro fittings & installation to purchase cost for new pipe

o KF – Annual fixed charges including maintenance, expressed as fraction of initial cost for completely installed pipe

• Optimum Economic Pipe Diameter:

• The total annual cost for piping system & pumping for turbulent flow is:

2.84 0.84 0.16 n

CT = 0.273qf ρ µ K(1+J)Hy + B’ + (1+F)XDi KF

4.84

Di E

st

The optimum pipe diameter is obtained by taking the 1 derivative and equating to zero: i.e.

2.84 0.84 0.16 1/(4.84+n)

Di,o = 1.32qf ρ µ K(1+J)Hy

n(1 + F)XEKF

• The total annual cost for piping system & pumping for viscous flow is:

2 n

C’pumping = 0.024qf µc K(1+J)Hy + B’ + (1+F)XDi KF

4

Di E

st

The optimum pipe diameter is obtained by taking the 1 derivative and equating to zero: i.e.

2 1/(4+n)

Di,o = 0.096 qf µcK(1+J)Hy

n(1 + F)XEKF

LINEAR PROGRAMMING

• Most widely used optimization technique, perhaps the most effective

• The word programming does not refer to computer programming,

• But it means optimization (also true for non-linear and integer programming)

• Both objective function and constraints are linear (problems)

• A mathematical technique for determining optimum conditions for allocation of resources & operating capabilities to attain a

definite objective.

• Also useful for analysis of alternative uses of resources or alternative objectives

Objective: usually to maximize profit and minimize cost

Objective function: mathematical function

Decision variables – design variables

Constraints – limitations on the values of decision variables

9

ChE 522: Chemical Engineering Plant Design and Economics Lecture Notes

Common Constraints:

1. use of raw materials 3. time availability

2. man-hours (labor) 4. budget

RATIONAL APPROACH TO PROBLEMS INVOLVING LP

1. A systematic description of the objective

2. A systematic description of the limitations or constraints

3. A clear knowledge of decision variables

4. Combination of the constraints conditions and the objective function to choose the best result out of many possibilities.

Mathematical Formulation:

Decision variables: Let x1 = units of product 1 ; xn – units of product n

x2 = units of product 2

Objective function: Max profit: P = a1x1 + a2x2 + …. + abxn

Subject to: (constraints)

a11x1 + a12x2 + . … + anxn (≤,≥, = ) b1

a21x1 + a22x2 + …. + a2nxn (≤,≥, = ) b2

a1 , a2 ……………..xn ≥ 0

v Availability of RM, man hours availability, non-negativity constraints, others

Methods:

1. Graphical – applies only up to 2 decision variables only

2. Simplex – Gauss Jordan

Applies to at least 2 or more decision variables.

SIMPLEX ALGORITHM:

Optimality condition:

Maximization: the entering variable is the non-basic variable having the most negative coefficient in the Z equation.

Minimization: the entering variable is the non-basic variable having the most positive coefficient in the equation.

• A tie between 2 non-basic equation variables maybe broken arbitrarily

• Where all the left hand side coefficients of the Z-equation are non-negative, the optimum is reached.

Feasibility condition: The leaving variable is the variable corresponding to the smallest ratio of the current values of the basic

variables to the positive constraint coefficients of the entering variable.

A tie may be broken arbitrarily.

Problem 3:

Two products X & Y, both require processing time on machines 1 & 2. Machine 1 has 200 hours available, and machine 2 has 400 hrs

available. Product X requires 1 hour on machine 1 and 4 hours on machine 2. Product Y requires 1 hour on machine 1 and 1 hour on

machine 2. Each unit of product X yields $10 profit and $5 for product Y. Determine the number of units of products X & Y that

should be produced to give the maximum profit. Solve by simplex method.

Soln:

Decision variables: let X – be the no. of units of product X

Y – be the no. of units of product Y

Objective function: Max profit (P), P = 10X + 5 Y

Subject to: constraints

a) X + Y ≤ 200 b) 4X + Y ≤ 400 c) X, Y > 0

Standard form:

Max. P = 10X + 5 Y + 0S1 + 0S2 or P – 10X – 5Y = 0

ST: X + Y + S1 +0S2 = 200 or X + Y + S1 = 200

4X + Y + 0S1 + S2 = 400 or 4X + Y + S2 = 400

X, Y > 0

Where S – represent slack variable to convert the inequality to an equality

10

ChE 522: Chemical Engineering Plant Design and Economics Lecture Notes

Initial tableau

Basic X Y S 1 S 2 RHS

P -10 -5 0 0 0

S 1 1 1 1 0 200 200/1 = 200

S 2 4 1 0 1 400 400/4 = 100

Entering variable (EV): the non-basic variable w/ most negative sign (X)

Leaving variable (LV): the variable corresponding to the most (+) of the entering variable.

nd st

2 tableau (1 iteration)

New R3 = Old R3 ÷ 4 ; New R1 = New R3 x 10 + old R1

New R2 = New R3 x (-1) + old R2

Decision Variables Slack Variables

Basic X Y S 1 S 2 RHS

P 0 -5/2 0 5/2 1000

S 1 0 ¾ 1 -1/4 100 (100) / (3/4) = 133.33

X 1

¼ 0 ¼ 100 (100) / (1/4) = 400

EV: Y LV: S1 Pivot pt: (¾) LR = 133.33

rd nd

3 tableau (2 iteration)

New R2 = Old R2 x (4/3) ; New R1 = New R2 x (5/2) + old R1

New R3 = New R2 x (-1/4) + old R3

Decision Variables Slack Variables Optimum

Basic X Y S 1 S 2 RHS Answers:

P 0 0 10/3 5/3 4000/3 P = 1333.33

Y 0 1 4/3 -1/3 400/3 Y = 133.33

X 1

0 -1/3 1/3 200/3 X = 66.67

Problem 4: Carnell Electronics Inc., produces two models of electronic calculators, the X15 & Y10. The company has been extremely

surprised by the demand for its products. Until an expansion program is completed, the company will be able to sell as many of

either model as it can produce. Production capacity is limited in three departments: soldering, assembly and molding. Each month,

the company has 50,000 minutes of molding capacity. Either model requires 1 min of molding time. Model X15 requires 5 min to

assemble, Y10 requires 2.5 min. The company has 200,000 minutes of assembly time each month. Soldering operations require 3 min

on the X15 and 8 min on the Y10. The company has 300,000 minutes of soldering time each month. Model X15 contributes $10 to

profit, $8 for Y10. Determine the maximum profit. Solve graphically and by simplex method.

Soln:

Decision variables: let X1 = no. of units of model X15 produced

X2 = no. of units of model Y10 produced

Objective function: maximize profit:

Constraints: 1) molding capacity (50,000 min: 1 min for both)

2) assembly time (200,000 min) X15 = 5 min; Y10 = 2.5 min

3) soldering time (300,000 min) X15 = 3 min; Y10 = 8 min

4) X1, X2 > 0

11

ChE 522: Chemical Engineering Plant Design and Economics Lecture Notes

Objective function: max P; P = $10(X1) + $8(X2) ; P = 10X1 + 8X2

Subject to: 1) X1(1min/unit) + X2(1 min/unit) ≤ 50,000 ; X1 + X2 ≤ 50,000

2) X1(5min/unit)+X2(2.5 min/unit) ≤ 200,000 ; 5X1 + 2.5X2 ≤ 200,000

3) X1(3min/unit) + X2(8 min/unit) ≤ 300,000 ; 3X1 + 8X2 ≤ 300,000

4) X1, X2 > 0

a) graphical solution:

L1: X1 + X2 = 50,000 : pts: (0,50,000) & (50,000, 0)

L2: 5X1 + 2.5X2 = 200,000 : pts: (0, 80,000) & (40,000, 0)

L3: 3X1 + 8X2 = 300,000 : pts: (0, 37,500) & (100,000, 0)

X2 Points of intersection: extreme pts in feasible region

L 2 L1 & L3: X2 = 30,000, X1 = 20,000 & P = $440,000

L1 & L2: X2 = 20,000, X1 = 30,000 & P = $460,000

Optimum answers:

30,000 units of model X15

L1 20,000 units of model Y10

Shaded area: feasible region

L3

X1

b) By Simplex method:

Std from: max P = 10X1 + 8X2 or P – 10X1 – 8X2 = 0

ST: 1) 3X1 + 8X2 + S1 = 300,000

2) 5X1 + 2.5X2 + S2 = 200,000

3) X1 + X2 + S3 = 50,000

4) X1, X2 > 0

Initial tableau

Decision Variables Slack Variables

Basic X 1 X 2 S 1 S 2 S 3 RHS

P -10 -8 0 0 0 0

S 1 3 8 1 0 0 300,000 100,000

S 2 5 2.5 0 1 0 200,000 40,000

S 3 1 1 0 0 1 50,000 50,000

EV: X1 LV: S2 LR = 40,000

New R3 = Old R3 ÷ 5 ; New R1 = New R3 x 10 + Old R1

New R2 = New R3 x (-3) + Old R2 ; New R4 = New R3 x (-1) + Old R4

nd st

2 tableau : 1 iteration

Decision Variables Slack Variables

Basic X 1 X 2 S 1 S 2 S 3 RHS

P 0 -3 0 2 0 400,000

S 1 0 13/2 1 -3/5 0 180,000 26,692

X 1 1 1/2 0 1/5 0 40,000 80,000

S 3 0 1/2 0 -1/5 1 10,000 20,000

EV: X2 LV: S3 LR = 20,000

12

ChE 522: Chemical Engineering Plant Design and Economics Lecture Notes

New R4 = Old R4 x 2 ; New R3 = New R4 x (-1/2) + Old R3

New R2 = New R4 x (-6.5) + Old R2 ; New R1 = New R4 x 3 + Old R1

rd nd

3 tableau: 2 iteration

Decision Variables Slack Variables Optimum

Basic X 1 X 2 S 1 S 2 S 3 Answers

P 0 0 0 4/5 6 460,000

S 1 0 0 1 2/5 -13 50,000

X 1 1 0 0 2/5 -1 30,000

X 2 0 1 0 -2/5 2 20,000

Problem 5:

Catchem is a supplier of special chemicals to petroleum & chemical industries in the San Francisco Bay area of Northern California.

Martha Sellmore, sales manager of Calchem is trying to determine how the typical salesperson should divide time between

petroleum (PIC) & chemical industry customers (CIC). The ff information seems pertinent to the sales manager.

PIC CIC

Ave profit per sales call $500 $200

Hours of salesperson’s time req’d/ sales call 8 hrs 8/3 hrs

Ave. entertainment cost/sales call $40 $30

If less than 8 hours per day per salesperson is required for sales calls, other administrative duties can be performed to fill out the

day. A maximum of $60 per sales person per day fro entertainment cost is allowed under company policy. How many daily sales

calls to petroleum and chemical industry customers should each salesperson make, on the average, to maximize profits? Solve

graphically.

Problem 6:

Carnell Electronics Inc., produces two models of electronic calculators, the X15 & Y10. The company has been extremely surprised by

the demand for its products. Until an expansion program is completed, the company will be able to sell as many of either model as

it can produce. Production capacity is limited in three departments: soldering, assembly and molding. Each month, the company

has 50,000 minutes of molding capacity. Either model requires 1 min of molding time. Model X15 requires 5 min to assemble, Y10

requires 2.5 min. The company has 200,000 minutes of assembly time each month. Soldering operations require 3 min on the X15

and 8 min on the Y10. The company has 300,000 minutes of soldering time each month. Model X15 contributes $10 to profit, $8 for

Y10. Determine the maximum profit. Solve the problem graphically and by simplex method.

Problem 7:

A merchant plans to sell two models of computers at costs of $500 & $800 respectively. The $500 model yields a profit of $45 and

the $800 model yields a profit of $50. The merchant estimates that the total monthly demand will not exceed 250 units. Find the

no. of units of each model that should be in stock in order to maximize profit. Assume that the merchant does not want to invest

more than $140,000 in computer inventory.

Problem 8:

A fruit juice company makes two special drinks by blending apple and pineapple juice. The first drink uses 30% of apple juice and

nd

70% of pineapple while the 2 drink uses 60% of apple and 40% of pineapple. There are 1000 liters of apple and 1500 liters of

st nd

pineapple juice available. If the profit for the 1 drink is $0.60/L and that of the 2 drink is $0.50/L, find the amount in liters of each

drink that should be produced in order to maximize profit.

Problem 9:

Maximize Z = 3x1 + 4x2

Subject to: 2x1 + 5x2 ≤ ; 4x1 + 3x2 ≤ 12 ; x1, x2 ≥ 0.

Problem 10:

A chemist sells four scents of perfume. Each of these scents contains three types of perfume in the ff proportions:

Type of Perfume

At present, the quantities on hand are: 220 oz of type K, 240 oz of type

Scent K L M L, and 110 oz of type M. No additional amounts can e obtained during

the coming week. She sells at $1.95/oz of scent A, $1.90/oz of scent B,

$1.70/oz of scent C and $1.45/oz of scent D. The demand for these 13

scents during the coming week is

ChE 522: Chemical Engineering Plant Design and Economics Lecture Notes

A 20% 50% 30%

B 35% 40% 25%

C 50% 30% 20%

D 60% 25% 15%

expected to be such that the entire supply of whatever product mix is prepared can be sold. What are the constraints and the

objective function which should be taken into consideration determining many ounces of each scent should be prepared?

Problem 11:

A production facility is being used to produce three different products, x1, x2 and x3. Each of these products requires a known

number of employee-hours and machine-hours for production such that: Product x1 requires 10 employee-hours and 15 machine –

hours per unit. Product x2 requires 25 employee-hours and 10 machine-hours per unit. Product x3 requires 20 employee-hours and

10 machine-hours per unit. The profit per unit is $5 for x1, $10 for x2, and $12 for x3. Over the base production period under

consideration, a total of 300 of employee-hours and a total of 200 machine-hours are available. With the special restriction that all

employee-hours are to be used, what mix products will maximize the profits?

Problem 12: A confectioner manufacture two kinds of candy bars: Ergies (packed w/ energy for kids) and Nergies (for weight

watchers). Ergies sell at a profit of $0.50/box. The candy is processed in three main operations: blending, cooking, and packaging.

Each box of Ergies candy requires 1 min blending, 5 min for cooking and 3 min for packing while each box of Nergies requires, 2 min

blending, 4 min cooking and 1 min for packing. During each production run, the blending equipment is available for a maximum of

14 machine hours, the cooking equipment for at most 40 machine hours and packaging equipment for at most 15 machine hours. If

each machine can be allocated to the making of either type of candy at all times that it is available for production, determine how

many boxes of each kind of candy the confectioner should make to realize the maximum profit. Solve graphically and by simplex

method.

Soln:

Decision variables: Let X1 – boxes of Ergies & X2 – boxes of Nergies

Objective function: maximize profit, P; P = 0.5X1 + 0.60X2

Subject to constraints: a) X1 + 2X2 ≤ 840 → L1

b) 5X1 + 4X2 ≤ 2400 → L2

c) 3X1 + 2X2 ≤ 900 → L3

d) X1, X2 > 0

a) Graphical soln: L1 pts: (0, 420) & (840, 0)

L2 pts: (0,600) & (480,0)

L3 pts: (0,900) & (300,0)

From graph: optimum condition is the intersection of L1 & L3

From (1) X1 = 840 – 242

in (3) 3[840-2X2] + X2 = 900

X2 = 324 boxes of Nergies (0,900) L3

X1 = 192 boxes of Ergies

P = 0.5(192) + 0.6(324) = $290.4

L3 L2 L1

(300,0) (480,0) (840,0)

b) by simplex method:

Initial Decision Variables Slack Variables

14

ChE 522: Chemical Engineering Plant Design and Economics Lecture Notes

Basic X 1 X 2 S 1 S 2 S 3 RHS

P -0.5 -0.6 0 0 0 0

S 1 1 2 1 0 0 840

S 2 5 4 0 1 0 2,400

S 3 3 1 0 0 1 900

RHS

Basic X 1 X 2 S 1 S 2 S 3

P 1.3 0 0 0 0.6 540 NR1 = NR4(0.6)+OR1

S 1 -5 0 1 0 -2 -960 NR2 = NR4(-2) + OR2

S 2 -7 0 0 1 -4 -1,200

X 2 3 1 0 0 1 900 NR3 = NR4(-4) + OR3

NR4 = OR4

Basic X 1 X 2 S 1 S 2 S 3 Answers

P 0 0 -0.26 0 1.12 290.4

X 1 1 0 -1/5 0 -0.4 192

S2 0 0 -1.4 1 -8.2 144

X 2 0 1 0.6 0 2.2 324 NR1 = NR2(-1.3)+OR1

NR2 = OR2 / (-5)

15

- em8779-e[1]Uploaded byRobinSingla
- BESO3DUploaded byRachel Chang
- SRS-298Uploaded byCarmel Buniel Sabado
- 20.IJASRAPR201720Uploaded byTJPRC Publications
- Paper 016Uploaded byDaris Purnomo Jati
- dataleakagedetection123-131216094254-phpapp01Uploaded byQ-mar Nair
- MATH5.MODULE-BODY (1).docxUploaded byTin Robiso
- Scheduling-Optimisation-of-Chemical-Process-Plant.pdfUploaded byPierre Castillo Abril
- Linear Programming_problems_EN_29-5-2012.pdfUploaded byVinkal Prashar
- Linear Programming ExercisesUploaded bykamran
- Optimization HandoutUploaded byKaseh Dhaah
- Linear-programming.pptxUploaded byHassan Khan
- 2-78-1424077448-3. management - OPTIMIZING CEMENT DISTRIBUTION IN THE NIGERIAN CEMENT MANUFACTURING- Dr Nwekpa K C (1).pdfUploaded byParr
- EE16Uploaded byMohit
- WCSMO4 FAIPA SAND Herskovits Mappa Juillen 2001Uploaded byjuillenl
- Lot-sizing in flow-shop with energy consideration for sustainable manufacturing systemsUploaded byzawadzahin
- Us 8082167Uploaded bycajimenezb8872
- Fuzzy Linear Programs With Trapezoidal Fuzzy NumbersUploaded byGanesan Kandasamy
- Solving a Gas-lift Optimization Problem by Dynamic ProgrammingUploaded byCehan
- 8 Chapter 3 Graphical MethodUploaded byRoja sundaram
- A Review Paper on Classical Techniques to Solve Optimal Power FlowUploaded byInternational Journal for Scientific Research and Development - IJSRD
- CIRRELT-2016-30Uploaded byMichel Rodrigo
- LeNy Pappas ATC10 FixSchedulingUploaded byAdrian Dinca
- Transportation ProblemUploaded bypuneeth h
- PesaranUploaded byArthuro Bandini
- OptimizationUploaded bySebastián Cáceres G
- Lesson03GraphicalMethodForSolvingLPPUploaded byToshal Bootna
- daskinsnyderandberger_geradbook_Uploaded byShivangi Kumar
- daskinsnyderandberger_geradbook_.pdfUploaded byShivangi Kumar
- Linear-Programming-Simplex-Method.pdfUploaded byGeorgina Francheska Ramirez

- bsc it.pdfUploaded byKiran Murmu
- MBA-Linear Programming.pptUploaded byexams_sbs
- 08 LinearUploaded bychilledkarthik
- MB0048 (3)-1 Operation ResearchUploaded byRaushan Ratnesh
- Sem 3Uploaded byKunal Berde
- Mathematics_2019-20AB.pdfUploaded bySandeep thammana
- Search Methods for Finding Nash Equilibrium_paperUploaded bywalrod22
- Airfoil Aerodynamic OptimizationUploaded byihv69
- Course+Syllabi+MinEUploaded byabdullah.alanani
- AppEUploaded byArvind Shukla
- Form 5 Chapter 10 Linear ProgramingUploaded byJenn
- Study GuideUploaded byShaoPuYu
- Natural Gas Infrastructure Design with an Operational Perspective.pdfUploaded byjvenancio
- Linear ProgrammingUploaded bygvspavan
- an inquiry based approach for teaching students to formulate linear programming models.pdfUploaded byPaulAyvazian
- Herbert Simon Model of Decision MakingUploaded bySachin Raikar
- Chapter 1 Linear ProgrammingUploaded bymasyati
- Geometric PatternsUploaded bydudi
- Koller Friedman Probabilistic Graphical ModelsUploaded by정민수
- Integer ProgrammingUploaded byDamar Saloka A
- ReviewQA_9Uploaded byZainol Azhar Ali
- Exercise 5 (Chap 5)Uploaded bypeikee
- Business Analytics by James Evans TOCUploaded byArchie
- Project Time–Cost Analysis Under Generalised Precedence RelationsUploaded byVitor Feitoza
- NPTEL Gomory Cut Algorithm Course Linear Programming ProblemsUploaded bySaid
- Case Study on Profit Planning of Textile Industry Using Linear Programming ApproachUploaded byVamshee Krishna
- math 1010 linear programming project-2Uploaded byapi-267007020
- OR QP MCAUploaded bymytri84
- Suggested Answer Dec 2015Uploaded bychatur8407
- K.U.K-M.B.A 2nd Sem SyllabusUploaded byvivekatri

## Much more than documents.

Discover everything Scribd has to offer, including books and audiobooks from major publishers.

Cancel anytime.