See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/308763228 Application of Linear Programming in Mine Systems Optimization Article · September 2016 CITATIONS READS 0 972 1 author: Abinash Swain National Institute of Technology Rourkela 8 PUBLICATIONS 0 CITATIONS SEE PROFILE Some of the authors of this publication are also working on these related projects: Developing Risk Map w.r.t Strata Management and figuring out Causes of Failure Trend for effective changes in Strata Support Mechanism. View project All content following this page was uploaded by Abinash Swain on 30 September 2016. The user has requested enhancement of the downloaded file. Application of Linear Programming Problem for Mine System Optimization Report for: Mine Systems Engineering (MN 624) Guided by- Prof. A K Gorai, NIT Rourkela Submitted by: Abinash Swain, Roll no- 113MN0495 Date- 28.09.2016 1|Page CONTENTS Item no Name Page no. 1.0 Introduction 2.0 Areas of Application of LPP in mining Industries 4 2.1 Blending of Ores and Production Scheduling 4 2.2 Transportation Problems 5 2.3 Assignment Problems 7 2.4 Inventory Management and Optimization 9 2.5 Queuing optimization in Opencast Mines 12 2.6 Project Management 13 2.7 Project Crashing using Optimization techniques 16 3.0 Conclusion 17 2|Page 3 1.0 Introduction: Linear Programming methods and Operation Research techniques help us in perhaps all industry oriented work to optimize our system. Application of it in mining industry is vital. Uncertainty in demand in the market, continuously varying market conditions is the key problem for which optimization is required. Developed during the World War II in England, operation research has emerged as a vital tool now-a-days. It makes us take decisions scientifically giving best ever result for our profit optimization. Different mathematical operations and iteration techniques are applied in doing so. So many algorithms are developed in recent years for application in many fields to optimize our requirements. There are different stages of linear programming. First is the formulation of the liner programming problem or model. So, converting field real life data in to mathematical formulations is the important part. Then we can use several solving methods developed till date to get the best ever desirable result. By doing so we can eliminate probable loss in the business by beforehand taking measures. Hence linear programming problems helps mining industry in optimizing various things and thereby reducing the cost of production and increasing the profit. In mining industry the problems that can be solved by linear programming models are production scheduling according to market fluctuation, production from different parts of mine with varying quality for optimum blending parameters, transportation problem to different customers, Assigning various tasks to different group of people having different capabilities, Optimizing Inventory by calculating Order quantity, Reorder point etc. One of the major applications is the area of Project management. In mining it generally takes a lot of time to start production as it goes through various types of projects interrelated to each one. So estimating duration of project completion, allocating more resources to complete it in less time by crashing of project, finding critical path for optimization plays a pivotal role in optimizing overall profit and minimizing duration of completion keeping an eye on varying market conditions and resource availability. Problems are solved using different mathematical models. Algorithms are used for each type of model to solve it. May it be graphical or analytical modelling, an optimized point is reached among input parameters. Various Linear programming soft wares are now-a-days used for efficient planning in mining industries. 3|Page 2.0 Areas of Application of LPP in Mining Industries: Blending of Ores and Production Scheduling Transportation Problems Assignment Problems Inventory Management and Optimization Queuing optimization in Opencast mines Project management Project Crashing Management and many more. 2.1 Blending of Ores and Production Scheduling: We know the quality of ore in a single mine varies from place to place. But from a producer point of view, we have to maintain a certain fixed quality of ore to be supplied to the customer. We can’t give high quality ore on a day and low quality on the next day. Hence we apply the principle of Blending. By blending of ores of different qualities we can produce, every time, an ore of fixed parameters of ingredients. For this purpose we need production scheduling and optimization in blending process. What amount of different grades of ores will be mixed to produce the required quality of ore can be decided by using optimization technique through Linear Programming applications. As good quality ore gives us more money and low quality earns us less, hence we have to optimize our quality and hence maximizing the profit avoiding future losses. For an example: A mine makes three Products of ore of different quality by blending ores of different proportions. It has two types of ore called Ore A and Ore B to get three blend products of Product-1, Product-2 and Product-3. For making Product-1, it needs to blend Ores A and B in the proportion of 4:6. To make Product2, it needs to blend ores A and B in the proportion of 4:8, similarly for Product-3 it needs the ratio of 3:8. The selling price of Product 1, 2 and 3 per ton are 30$, 50$ and 35$ respectively. So how can we decide that what product mix/blend gives us maximum profit so that we can optimize our profit. Given that the maximum production of Ores A and B can be 800 tons and 2000 tons respectively. 4|Page Approach using Linear Programming: Let’s assume that we make products 1, 2 and 3 as X1,X2,X3 tons respectively. So our Objective function is a cost function Z. It can be written as: Z= 30X1+50X2+35X3 We have to maximize it to get maximum profit. But we have certain Constraints to solve the problem. These constraints can be written as: 4X1+4X2+3X3 <= 800 6X1+8X2+8X3 <=2000 This is a simple linear programming problem. We can solve it using Simplex method to solve it easily and can get the tones of product-1, 2 and 3 giving maximum profit and hence thereby we can plan our schedule for production of different types of ores on day to day basis. 2.2 Transportation Problem: A mine produces ores to be transported to consumers. For example a coal mine sends its coal produced to different power plants or steel plant to be used as fuel or coking coal. A single mining company usually does not have a single customer. So it sends varying quantity of ores to different plants from its different operating mines. As there is a large spending in transportation of ore is involved in the process, the mining company has to optimize the quantity of ore to be sent to different customers from different mines located at different locations and involving varying transportation cost. We can use Linear Programming technique to maximize our profit in transportation by optimizing the cost involved. For an example: A mining companies has three mines named mine-1,mine-2 and mine-2 respectively located at different places and produce coal of 150t,175t and 275t respectively. Similarly, there are three power plants to where these coals are to get transported. Demand at those plants named plant-A, plant-B and plant-C are 200t, 100t and 300t respectively. The cost of transportation of coal to these plants from different mines are given as: 5|Page So we have to optimize the quantity of coal to be transported from mines to different plants so as to minimize the cost involved in transportation. Also we have to make sure that all the produced coal is being transported and all the demand in the power plants are properly met. Approach using Linear Programming: So our objective function is a cost function Z and we have to minimize it. Let’s assume the coal transported from mine-1 to plant-A is notated by X1A and other notations are similar. Objective function is given by: Minimize Z = 6x1A + 8x1B + 10x1C + 7x2A + 11x2B + 11x2C + 4x3A + 5x3B + 12x3C But the constraints are given as: X1A + X1B+ X1C 150 X2A+ X2B + X2C 175 X3A+ X3B + X3C 275 X1A + X2A + X3A = 200 X1B + X2B + X3B = 100 X1C + X2C + X3C = 300, XIJ 0 This problem in linear programming can be solved by constructing Initial tableau and Optimal Solution tableau. Initial feasible solution can be found out by methods like: 6|Page North-West corner method Minimum Cell Cost method Vogel’s Approximation method And after getting initial feasible solution we have to check the solution is optimal or not. For that we can use Modified Distributed method (MODI). It is a modified method of Stepping Stone method. It determines a tableau is optimal or not. Also it tells us which non-basic variable to be chosen first to be replaced. 2.3 Assignment Problems: In a particular mine, the man power is not uniformly capable to do different types of works. Someone can do a certain job in less time due to his efficiency and interest, he may not be able to do another type of work with that amount of interest and dedication. Hence we have to optimize the distribution of work to the man-power available so as to minimize the time required to finish the job and hence indirectly increasing productivity and profitability of the mine. Hence use of Linear Programming and its algorithms helps immensely to assign jobs to different groups of man power and hence utilizing Human resources efficiently. For an Example: A production supervisor is considering how he should assign the four jobs that are to be performed, to four of the workers. He wants to assign the jobs to the workers such that the aggregate time to perform the jobs is the least. Based on previous experience, he has the information on the time taken by the four workers in performing these jobs, as given in table: Worker Job A B C D 1 45 40 51 67 2 57 42 63 55 3 49 52 48 64 4 41 45 60 55 7|Page So we can use LPP application to find out which work should be assigned to which man to minimize the time and maximizing the productivity. Approach using Linear Programming: Here Hungarian Assignment model can be implemented. Different Steps are: Step 1: Locate the smallest cost element in each row of the cost table. Now subtract this smallest element from each element in that row. Step 2: In this reduced cost table obtained, consider each column and locate the smallest element in it. Subtract the smallest value from every other entry in the column. As a result, there would be at least one zero in each of the rows and columns of the second reduced cost table. Step 3: Draw the minimum number of horizontal and vertical lines that are required to cover all the ‘Zero’. If the number of lines drawn is equal to n (the number of rows/columns) the solution is optimal, and proceed to step 6. Step 4: Select the smallest uncovered (by the lines) cost element. Subtract this element from all uncovered elements including itself and add this element to each value located at the intersection of any two lines. Step 5: Repeat steps 3 and 4 until an optimal solution is obtained. By following this LPP algorithm, we can reach at a solution table given as: Worker Job A B C D 1 5 0 11 14 2 15 0 21 0 3 1 4 0 3 4 0 4 19 1 8|Page 2.4 Inventory Management and Optimization: Inventory management is a vital part in any mining industry as far as cost and efficiency are concerned. With fluctuations in the demand and price level in the market for various reasons, our order, re-order, production level etc. are hampered to a large extent. If demand is too less, we just can’t keep on producing at a regular rate, as our inventory or storage cost will increase. Similarly, we can’t keep on ordering our required goods at that rate if demand deceases. Similarly, if demand increases abruptly we have to make sure that we do have required stock to meet the growing demand to get maximum profit. So inventory management and optimization helps us in different ways: • demand information is distorted as it moves away from the end-use customer • higher safety stock inventories to are stored to compensate • Seasonal or cyclical demand • Inventory provides independence from vendors • Take advantage of price discounts • Inventory provides independence between stages and avoids work stoppages Various inventory costs are given below. We have to minimize the total cost of inventory by optimizing our Economic Order Quantity (EOQ), Re-order point or time to re-order. Costs are: • Carrying cost- cost of holding an item in inventory • Ordering cost- cost of replenishing inventory • Shortage cost- temporary or permanent loss of sales when demand cannot be met. Benefits of Inventory management are summarized as: • Hedge against uncertain demand • Hedge against uncertain supply 9|Page • Economize on ordering costs • Smoothing There is always a trade-off between holding cost and ordering cost. If we order in large amount, our ordering cost gets lowered, nut of we don’t have that amount of demand our holding cost increases. Hence, we have to optimize our cost to minimize it and increase profitability. But demand is rarely predictable. As our demand varies time to time. Hence we can use some Safety Stocks in our inventory to minimize loss at peak demand hours. Also the time lag between ordering and receipt should be optimized. By considering all these parameters we can use linear programming and models to maximize our profit. Models can be used are: Deterministic model Probabilistic model 10 | P a g e 11 | P a g e For an Example : Assuming a mining company that faces demand for 5,000 te explosive per year, and that it costs Rs. 15,000 to have the explosive shipped to the mines. Holding cost is estimated at Rs. 500 per te per year. Estimating the standard deviation of daily demand to be d = 6. When should we re-order if you want to be 95% sure you don’t run out of explosive? Approach using Operation Research Technique: Since the expected yearly demand is 5000 te, the expected demand over the lead time is 5000(10/365) = 137 te. The z-value corresponding to a service level of 0.95 is 1.65. So ROP 137 1.65 10(36) 168 Order 548 te explosives when the inventory level drops to 168 te. 2.5 Queuing Optimization in Open-Cast Operations: Queuing theory helps immensely to optimize the dumper-shovel combination operation by increasing efficiency in production, decreasing cycle time and increasing profit indirectly. By using previous data in this technique we can find out average no. of dumpers being in the queue regularly, cycle time, waiting time etc. in a probabilistic form. So that we can take remedial steps to increase productivity by increasing paths, servers or Shovels, dumpers etc. by proper management using mathematical tools and operation research. For an Example: Suppose there are arrival of dumpers is 50/hour and there are certain no. of servers or Shovels in the mine serving 40/hour. We have to decide optimal no. of servers required to maximize profit and given that waiting time for dumpers in the system is 50$ per hour and service cost is 30$ per hour. Approach using Operation Research technique: Here we can find out the total cost by doing parametric analysis for different server numbers/ Shovel numbers. And from that the server number giving minimum cost can be taken in to consideration. 12 | P a g e We can find cost for waiting of dumpers by multiplying cost/waiting /hour multiplied by the average waiting time per dumper (Wq *50) and similarly (S *30) will give the corresponding service cost for each value of ‘S’. Hence by taking various values of ‘S’, we can analyze the value giving minimum cost. 2.6 Project Management Use of operation research and LPP is of paramount importance in Project management in mining industries. As we know mining activity does not commence at once, but it takes enough time for production after completion of certain number of project activities. Also each activity possesses certain constraints to get executed. Hence optimizing project completion time and scheduling is important as our aim is to start producing as early as possible. Project managers rely on PERT/CPM to help them answer questions such as: What is the total time to complete the project? What are the scheduled start and finish dates for each specific activity? Which activities are critical and must be completed exactly as scheduled to keep the project on schedule? How long can noncritical activities be delayed before they cause an increase in the project completion time? 13 | P a g e For an Example: A mining company needs tasks named A,B,C,D and E to complete in certain interval as a part of a project. Hence we can find the expected time of completion of project, critical path, critical tasks etc. to optimize our project work and minimize the duration. It is given: Task Immediate prerequisite tasks Effort A None 9 B A 5 C A 7 D B,C 11 E D 8 (Weeks) Approach using Operation Research Techniques: 14 | P a g e Blue circle gives the critical tasks that constitute the critical path. Or the path giving the longest duration of the project to get completed. This is the strategic path as we can minimize the project duration by changing the duration of critical tasks so as to maximize our productivity by minimizing the duration of project. The critical path method analyses the precedence of activities to predict the total project duration. The method calculates which sequence activities (path) has the least amount of flexibility. Different Steps for PERT network analysis: Step 1: Make a forward pass through the network as follows: For each of these activities, i, compute: • Earliest Start (ES) Time = the maximum of all earliest finish times for all its immediate predecessors. (For node “START”, this is 0.) • ESi= Maximum (EFj) for all immediate proceeding activities j. • Earliest Finish (EF) Time = (Earliest Start Time) + (Time to complete activity i). • EFi= ESi+ ti Step 2: Make a backwards pass through the network as follows: Move sequentially backwards from the last node, “FINISH” to its immediate predecessors, etc. At a given node, j, consider all activities immediately following it and compute: • Latest Finish (LF) Time = the minimum of the latest start times for all activities that immediately follow j. (For node “FINISH”, this is the project completion time.) • LFj= Minimum (LSi) for all immediate following activities i. • Latest Start (LS) Time = (Latest Finish Time) - (Time to complete activity j). • LSj= LFj - tj Step 3: Calculate the slack time for each activity by: • Slack = (Latest Start) - (Earliest Start) or (Latest Finish) - (Earliest Finish). • A critical path is a path of activities, from node “START” to “FINISH”, with 0 slack times. 15 | P a g e 2.7 Project Crashing using Optimization Technique: As we saw the critical path in the PERT network analysis contains the critical tasks that contribute to the maximum time for completion of the entire project. Hence to reduce project duration and increase profitability we have to crash certain critical tasks by providing more resources in to it. But obviously it will increase the cost of production. Hence we have to optimize our duration and cost for maximum productivity and profitability. Project crashing can be done only to certain allowable durations and limited resources. Hence use of LPP is vital in choosing crashing amount. For an example: Below diagram shows the network for finishing a certain project. It shows the expected time of completing individual tasks along with the maximum time that a task can be crashed using limited resources and rime. Approach using LPP technique: So, can we crash the entire time available to each task? The answer is NO. Because the path will not remain the critical path, hence our cost will increase for the project. So we can crash up to that amount so that we will never have loss due to crashing. Here we can crash all the time available because after that the path remains as critical path as 4+3 >5. But if the situation is like this shown below: 16 | P a g e We cannot crash all the weeks available as the path will not remain critical. In this way we can formulate mathematical models for crashing operations. Various Soft wares are available for executing so in mining industries. Below diagram shows the Time-Cost Tradeoff. 3.0 Conclusion: Hence, use of LPP in mining industries is of paramount importance. It helps industries to optimize its systems and increase productivity. Moreover, it helps in reducing cost and increasing profitability by optimization techniques. Use of LPP also helps to combat unforeseen conditions due to market fluctuations in demand and supply. Also we can utilize discount options and hence reducing total cost. Starting from the project management to inventory management and HR management, it has its larger role. 17 | P a g e 18 | P a g e View publication stats