Distillation Column - Report Example

Dr: Manolis Tomadakis Department: Chemical Engineering Florida Institute of Technology Address: 150 W Blvd, Melbourne, Florida, 32901 Date: 6 November 2014. Dear Dr. Tomadakis, Our group has come up with a unique sieve-plate distillation column for the separation of the ammonia-water mixture. This will enable the recovery of ammonia from the mixture. Thus, the feed flow rate (=0.28, 0.9985, =0.05) has been assigned. These values have enabled the group to come up with a design of the distillation column in line with Ponchon Savarit methods. Furthermore, the optimum reflux ratio was determined and placed at 1.02483 and the optimum reflux ratio to the minimum of 0.7392. Furthermore, the design set the height at 21.07m, the diameter at 3.95m and the tray spacing at 0.6m. The total cost of this design of distillation column is USD 21,113,329.98. Last but not least, we provide calculations on various parameters and costing, taking into consideration environmental impact of the plant, based on basic assumptions. Yours faithfully, Names. Summary Ammonia needed to be separated from a 22,000 aqueous solution through a sieve-plate distillation column. Tube heat exchangers and shell serve as condenser and reboiler, continuously operating for 350 days annually at 20 atm pressure. The feed goes in as a saturated liquid of 0.28 molarity fraction. The ultimate composition of the products is 0.9985 for the distillate product and 0.05 for the bottom product at a calculated efficiency of 46.91%. The steam had a temperature of 225 and pressure of 626.418 psia. The number of trays was 28 with 0.6m plate spacing for each. The column height and diameter were 21.07m and 3.95m respectively. A computer program, MATLAB, modelled the column conditions at various reflux ratios so as to find the solution to the Ponchon Savarit method yielding an optimum reflux ratio and Ropt of 1.02483 and 0.7392 respectively. Calculations of the total operating cost on MS-Excel gave USD 21,113,329.98. The group determined the annual cost and the other parameters on MS-Excel. The reflux number, referring to the number of real plates, and values obtained from MATLAB were used. Details on additional data and tables will be evaluated in the subsequent sections. Introduction Distillation column refers to a process of separation used to separate mixtures based on critical factors in the required process. In a chemical unit of ammonia, the recovery of ammonia from an aqueous solution being supplied at a flow rate, F = 22,000 kgmol/hr requires a fresh sieve-plate distillation column. The continuously running column will operate at a pressure of 20 atm with the feed entering as saturated liquid, having 0.28 xf molar fraction. The analogous compositions required are 0.9985 Xd for the distillate and 0.05Xw for the bottom products, the efficiency of the process being 0.4691. The optimum reflux ratio, determined to the minimum, was 1.02483. The heat steam cost would escalate the fixed cost with the utility pressure of steam being dependent on the determination of distillation column annual operating cost. In this report, the consequence of changing parameters (xd, xf, xw, Ps, F and R) and the annual cost will be given. Assumptions of the Design The column will be continuously operating at 20 atm with only 15 days maintenance break per year. Additionally, the feed will go into the column as saturated liquid. The new shell-and-tube heat exchangers give the plant the capability to withstand shell and tube pressures of 20 atm and 50 atm respectively. Based on these pressures and the expected flow rates and temperatures, preliminary design calculations indicated the application of Uc and Ur as the empirical correlations for heat transfer coefficient for each of the condensers and reboilers respectively. The cooling water at the cooling tower was provided at a temperature of 20 with an expected optimum operation reached if the water streaming back to the tower has its temperature at about 30. The pipelines were all well insulated. The pressure of saturated steam was 626.418 psia, this being high enough to ensure that the steam is not less than 40 times hotter than the mixture meant to be vaporized contained in the partial reboiler. Flooding in the column is prevented by allowing a maximum vapor velocity at the top, with its value being dependent on the densities of liquid and vapor on the top plate and on the plate spacing at 0.6m. The overall column efficiency was placed at 0.4691 as derived from O’Connell correlation based on volatility and viscosity. Heat losses and effects with regards to the column are negligible. Process Description Parametric study: MATLAB and MS-Excel programs were used in our calculations for designing a distillation column factoring in costing and parameters values. MATLAB was programmed to solve through the Ponchon-Savarit method. Nonetheless, this approach required that some factors such as the distillate composition, flow rate, column efficiency, distillate bottom composition and the feed mole fraction be input into the program. Inputting these values into MATLAB was programmed to give the Ropt with any value of reflux when the values of the parameters are changed. Running this program several times gives the Ropt/Rmin, this being the critical point for changing from decreasing to increasing the total annual cost. This change is referred to as Ropt for the given R value. The MATLAB program gives an output of the calculation of vapor and liquid fractions for ammonia at each phase or plate (X, Y) and temperature. Additionally, the program calculates the Qc, Qr, Vt and np values to find the solution for the other equations for diameter and costing, velocity and height in MATLAB or MS-Excel. For the input parameters, MATLAB was used to find Ropt, Qc, Qr, Np and Vt. Inputting Ts steams, Ps steams and densities for the ammonia-water mixture in MATLAB enabled their determination from NIST and steam table. The calculation of efficiency as based on volatility and viscosity yields 0.4961. The output from MATLAB is used on MS-Excel to determine the total annual cost, Ac, Ar, h, d, Vm and Vt using the appropriate equations. The results from MATLAB gave the Ropt for our design as 0.7392 and the ratio as 1.02483. These figures could change if different values for xd, xw and xf were used. Van der Waal’s equation was used to determine the maximum velocity on MS-Excel by taking the tray spacing as 0.6m and multiplying the diameter by 15% so as to improve on the safety factor and overdesign. The column height was determined by calculating the number of real plates on MATLAB and multiplying it with the tray spacing, adding 14 ft on top. Cost Estimation Fixed cost: The fixed cost attributed to the equipment used in the distillation columns will be determined by the costing of each column (ANCCOL), reboiler (ANCREB) and condenser (ANCCOND). This costing establishes the feasibility of installation of a new sieve plate distillation column. The fixed cost could be calculated using either MATLAB or MS-Excel. This was determined through solving of equations on each input parameters case using MS-Excel so as to establish the installation, purchase and operation costs of the equipment. With this regard, the Marshall and Swift index was taken as 1450/925 for 2012, then calculating the diameter using molar volume which is dependent on the R value. The summation of ANCCOL, ANCCOND and ANCREB gives the fixed cost, which is USD 272,199.312, given Ropt = 0.7392 and the ratio = 1.02483. The final cost of an installed column is computed from the value of diameter so as to calculate Cd$/plate = USD 12,211.30/plate. This is then multiplied by the number of real plates = 28 and the Marshall and Swift index = 1450/925 so to yield the total cost of the installed column = USD 535,977.00. Thereafter, the cost of piping, instrumentation and insulation is included to the cost found. This is done through the multiplication of the cost with (1 + 0.6) [range 0.6-0.9]. This gives USD 857,563.20. Finally, the fixed charges or annual cost of the column, ANCOOL, is calculated. This is done by multiplying the determined cost by 0.15, giving the cost as USD 128,634.48. The cost of the installed condenser is also established through calculations to determine the Ac value, these being equations with parameters earlier solved in MATLAB or MS-Excel. The first step involves calculating Uc using the Vn value in both equations in MATLAB and multiplying it by Xd. The Qc value = 41,691,000.9 Kcal/hr, had already been determined from MS-Excel for the given ratio. The final major factor needed is 25.1799782, which is pegged on the temperatures of the inlet and outlet cooling water, a value already computed from MS-Excel using the appropriate formulae as indicated in the Appendix section. Having these values, it became possible to determine Ac = 955.382760m2 from MS-Excel. Using this value in Cc gives the total cost of the installed column which is USD 140487.4382. The same procedure was followed to determine the annual cost of the column. Thus, the cost of the installed condenser was multiplied by 0.15, the Marshall and Swift index = 1450/925 and (1 + 0.6) to determine the fixed charges, ANCCOND = USD 52,853.65242. In the same way the condenser cost was determined, so was the cost of the installed reboiler also determined. First, Ar was solved taking the value of Qr from MATLAB as 64,441,461.9 Kcal/hr. Thereafter, the value of Ur was solved for from the A, B and C equations based on the pressure steam and Vn value, also determined from the MATLAB program. Subsequently, () is determined based on the steam temperature. This is done by adding 450 to highest temperature of the mixture. This was determined by drawing a straight line as determined by xw. With the determination of the value of Ar = 1,262.014126 m2, based on parameters earlier determined, Cr could be solved for, this being the cost of the reboiler, USD 241,114.49. This amount was then multiplied by 1.6 × 0.158(1450/925) so as to yield the annual cost of the reboiler, ANCREB = USD 90,711.18. Annual Operating cost: Critical to note when computing for the annual operating cost of the distillation column is the fact that it runs 24 hours a day for 350 days. The break of 15 days a year is meant for maintenance. Thus, the column operates 8,400 hours a year. From Peter and Timmerhaus Table 5, estimates on the cost for cooling and steam water for each of the condensers and the reboiler between January 2002 and January 2012 could be determined. Summing up the cost of heating steam and cooling water gives the annual operating cost = USD 20,841,130.66 for Ropt = 0.7392 and ratio = 1.02483. Computing the cost of cooling water on MS-Excel is done by dividing the condenser duty value, Qc = 41,691,000.9 kcal/hr by the average specific heat of water, this being the capacity of the given water at a given ambient temperature. This gives Gw = 4,169,896.86 kg/hr. Then, this would be multiplied by the cost incurred to cool a kg of water as provided for in the table as USD 0.000045 and 8,400 hrs/year. Therefore, the annual cost for cooling water would be determined after its multiplication with the price per kilogram and the operational time, giving ANCCW = USD 1,576,221.01. The cost of heating steam would be determined on MS-Excel by dividing the Qr value = 64,441,461.9 Kcal/hr by the latent heat = 1,689.8 KJ/Kg, which is deduced from the steam table using pressure steam or temperature, giving us Gst = 159,665.9444 Kg/hr. This would then be multiplied by the cost of steam and operational time. As derived in 1990, the cost of steam was found to be USD 3.7080/1000lb at the given the pressure steam. This has already been computed for in the calculations part of the report. Thus, in 1990, the cost of steam was USD 0.00814/kg. In 2014, this would be multiplied 1.68, giving us USD 0.013674/kg. This would then be multiplied by the duration of operation, 8,400 hours/year, and Gst so as to yield the annual heating steam cost for 2014, which is USD 19,264,909.65 ANCHS. Total annual cost: This value would be determined by summing up the annual operating cost and fixed cost at Ropt = 0.7392 and ratio = 1.02483. Thus, the total annual cost amounts to USD 21,113,329.98. Optimization Process: MS-Excel and MATLAB are critical programs that have helped in finding the optimum values for the design. The programming of MATLAB was meant to compute the total annual cost. The other calculations were calculated on MS-Excel so as to save on the time determining the R values for the given parameter cases, computing for Ropt at different values of R. MATLAB has the ability to calculate and determine Ropt for changing parameters like xw, xd, xf, Ps, F and Ropt/Rmin with all the equations required in the Ponchon Savarit methods and issues on costing. Ropt could be changed with the alteration of the value of any parameters. Eventually, all parameters are plotted against the total annual cost. Each of these cases will be explained in the subsequent sections. Reflux ratio: Given the observations made from this study and the values carefully calculated, it is appreciated that there is an inverse proportional relationship between the reflux ratio and the fixed ratio. Thus, increasing the reflux ratio leads to decrease in the fixed cost. However, increasing the reflux ratio increases the total operating cost. The lowest total annual cost given R = 0.7392 is Ropt. Thus, the total operating cost has a directly proportional relationship with the reflux ratio, implying that as reflux ratio increases, the total operating cost also increases. However, because of the increased demand for cooling water at the condenser, the reboiler and condenser area increases. When the fixed cost decreases, the reflux ratio increases. This inverse proportional relationship results from the number of real plates which will decrease as our calculations show. Feed flow rate, F: The MATLAB program aimed at establishing the relationship existing between changing the feed value against the annual cost, given that the other parameters are held constant as given. An increase in the feed flow rate leads to an increase in the fixed cost, total annual cost and operating cost. Other factors also affected in the same manner are the height and diameter of the column. Furthermore, the heat and area of the reboiler and condenser increase due to the fact that the more the feed, the more the heating and cooling needed. As the bottom and distillate flow increases, F increases. A decrease or increase of the feed flow rate has no effect on the number of trays and optimum reflux ratio. Pressure of steam: For as long as the pressure of the steam is increased, there is an increase in temperature. Because of the constant temperature of the surroundings, there is an increase in the heat lost to the surroundings. The changeability of the heat causes a reduction in reboiler efficiency. The area of the reboiler will decrease but that of the condenser will increase given that the pressure of the steam keeps rising. Additionally, with an increase in the pressure of the steam, the total annual cost increases, just as the annual operating cost also increases. However, fixed costs decrease. Feed Composition, Xf: An analysis of the calculations and values obtained from our study indicates that an increase in the feed composition results in an increase in the total annual cost, fixed cost and operating cost due to the increased cooling and heating requirements for the condenser and reboiler. Furthermore, with an increase in the feed composition, the height and diameter of the column increase. However, every feed composition has its unique Ropt. Thus, Ropt is not constant, same as the number of plates which could be altered. Distillate Composition, Xd: Altering the value of distillate composition affects other parameters, including vapor density at the column top. Increasing the distillate composition also increases the Ropt, together with an increase in the height and diameter of the column. Similarly, increasing the composition of the distillate will result in an increase in the total annual cost, including the operating and fixed costs. Bottom composition, Xw: Given the findings from our study, increasing the values of bottom composition results in an increase in the reboiler area and reboiler duty. In the same way, increasing the bottom composition causes a decrease in the diameter and height of the column and the number of real plates. Nonetheless, the fixed costs will keep decreasing due to an increase in the bottom composition. As long as the bottom composition is increased, the total annual cost decreases. Tray specification: The normal plate spacing ranges between 6 in. (0.15m) and 36 in. (0.9m). For column diameter greater than 1, the proposed plate spacing commonly ranges between 0.3 and 0.6m. Our design had a plate spacing of 0.6m. Normally, a tray of 24 in. (0.6m) would be considered as the minimum to make maintenance easy and also provide the optimum spacing for a wide array of conditions. Moreover, we increased the top tray spacing by 4 ft. so as to make it easy to remove entrained liquid. We added 10ft. spacing to the bottom tray to allow for bottoms surge capacity. For diameter measuring greater than 22 ft., it would be preferable to arrange at least two columns in series. Typical sieve trays have their columns with hole diameter of either ¼ in. or 3/16in. when carbon steel is used or 3mm (12 gauge) when stainless steel is used. These holes are distributed evenly and take up not less than 10% of the active area, this being the area between the outlet weir and the downcomer, this rarely going below 6%. They are placed on equilateral triangles, the pitch:diameter ratio ranging between 2 and 5, with 3.8 being the optimum. The design of these trays have tray thickness:hole diameter ratio ranging between 0.1 and 0.7 with between 10% and 20% of the entire area that a tray occupies taken by downcomers. Often, this area will be near 10%. In commercial columns, segmented vertical downcomer is the most widespread design. It is an inexpensive design to build, easy to install with an almost impossibility for installing incorrectly and could be customized to work with a wide array of liquid flow rates. Our chosen distillation column design is the four-pass tray. The flow patter was selected based on the height of the column meant for the four-pass design and the optimum liquid flow rate. The distance between the given hole centers, referred to as the hole pitch, should be 2.0 or greater with diameters normally ranging between 2.5 and 4.0. Within these specifications, the pitch could be selected so as to give the required number of active holes given the total hole area. Though equilateral square patterns are used, the triangular pattern is most preferred. Results: Eventually, the design for the distillation column designed for our plant has a diameter of 3.93768469m and a height of 21.0672m. At the optimum reflux ratio, Ropt = 1.02483, the column has its number of trays as 28, this determined after several runs on the MATLAB program. The optimum reflux of 0.7392 gives the value where change in position from decreasing to increasing occurs, referred to as the critical point. The selected tray spacing and liquid path length are 0.6m and 2.06ft respectively. The distillation column was designed to withstand pressure of 20 atm, operating at a feed flow rate of 22,000 kgmol/hr. The feed has an ammonia composition of 0.28, making the column capable of producing a distillate composition and bottom composition product of 0.9985 and 0.05 respectively. The condenser area is 955.3827604 m2 while the reboiler area is 1,262.014126 m2. The total annual cost sums up the amount of operating cost and fixed cost, thus USD 21,113,329.98. The Ponchon Savarit methods give: F = 22,000 kgmol/hr; Xw = 0.05; Xf = 0.28; Xd = 0.9985 Where; Xw refers to the composition of the liquid leaving the bottom of the tray; Xf is the feed composition; and Xd is the composition of the liquid leaving the condenser from the top of the column, otherwise the composition of vapor leaving the top tray. Overall column mass balance would be expressed as: F = D + W Where W and D refer to the bottom and distillate product streams respectively. The distillate flow rate could be computed from the value of the feed and the ratio of feed composition to bottom product composition. D = F × Xf = (22,000 kgmol/hr) × (0.28 – 0.05)/(0.9985 – 0.05) = 5,534.739 kgmol/hr. Then, the flow rate could be computed using the overall mass balance. Thus: W = F – D = (22,000 kgmol/hr) – (5534.739 kgmol/hr) = 16465.261 kgmol/hr The point A gives the molar enthalpy determined by connecting point Xd on the given equilibrium curve to vapor enthalpy curve at the top part of the chart. Similarly, point L is reached through a connection of point Xd to liquid enthalpy curve, as opposed to the vapor enthalpy curve. This occurs directly underneath point A. Points A and L form a line whose length, given the composition, equals the latent heat, thus the distillate latent heat. λd = AL = kcal/mol Thereafter, location of point ‘F’ follows, giving the feed molar enthalpy. This would be computed by first locating the feed vapor composition, this being the equilibrium vapor. It gives the composition of the liquid that enters the column coming from above the feed tray and that of the known liquid and feed: (Xw) / (Xd – Xw) The drawing of a horizontal line running from point Xf to the vapor curve yields Yf. Point ‘F’ could be determined in the same way points ‘A’ and ‘L’ were determined, starting from Xf instead of Xd, giving vapor enthalpy, Hf drawn to the vapor curve. To locate point Nm, the expression below applies: Nm = F + [(Xd – Xf)(Hf)] To determine the minimum possible reflux ratio, Rmin, the below expression applies: Rmin = (Nm – A) / λd Thereafter, R is computed as Rmin < R < 1.2Rmin To find out the length of line NA: NA = λd × R After the length between ‘A’ and ‘N’ is resolved, ‘N’ is determined by drawing a line from ‘A.’ Thus: N = A + NA The condenser removes heat from the distillation process which equals the distillate flow rate in the condenser multiplied by the total latent heat determined from the measurement of the distance between ‘L’ and ‘N.’ Hence: Qc = D × (1 + R)λd Drawing a line from Xw to the liquid enthalpy curve determines point B. A similar process yields point ‘M,’ thus: M = F + [(N – F)(Xf)] Total heat in the process, attributed to the reboiler, equals the bottom product flow rate multiplied by the total latent heat as determined from the distance from ‘M’ to ‘W’ λw = B – M = W × λw The value of Qr is determined knowing Yf and v – F)/(Yfv – Xf)] – Xw)/(Xf – Xd)] Technical, Environmental, Economical and Safety Considerations It would be appreciated that ammonia is corrosive and under high temperatures, it is potentially explosive. Since ammonia occurs in the form of liquid in the bottom product and distillate, there is no major concern about air emission by ammonia. Similarly, emission of the ammonia in vapor form is highly unlikely unless there is a catastrophic failure of the column or the condenser system, together with a leakage of the distillate stream. The rooms for storage or processing of ammonia should be ventilated adequately. Wet scrubbers would be ideal in the ventilation system to ensure safety when the product storage, distillation column or condenser is damaged. Furthermore, capture hoods play a vital role when one seeks to reclaim the ammonia that has leaked from the column, piping or storage. This being a mixture of water and ammonia only, any work safety effects and hazardous environmental effects would be solely attributed to ammonia. Generally, ammonia does not persist in air, soil or water, even though it is a vital nutrient for plants and bacteria. It does not accumulate in food chains. The wet scrubber system, together with the requisite fans, could cost between USD 70,000 and USD 215,000 as at 1995 in addition to the capital costs of the plant. This should be considered with reference to the possibility of the daily ammonia emissions hitting the limits set by the US Occupational Safety and Health Administration, OSHA. Considered a fire hazard, ammonia could easily explode upon ignition of the mixture of air and vapor; could explode when containers of ammonia are exposed to considerable heat; and emits poisonous fumes on decomposition upon heating. When working with ammonia, one needs to protect the eyes. The requirements for extinguishing ammonia fire include fine water spray, breathing apparatus and liquid tight protective clothing. Irrespective of the route used, exposure to ammonia is potentially harmful, specifically considered as being corrosive and toxic. Short term exposure causes burns to the throat and mouth, soft tissue irritation and swelling or watering of the eyes, sensitivity to light, cough, headache, confusion, sore throat and breathing difficulties. In the long run, exposure to ammonia could cause asthma, phlegm, persistent cough and wheezing. Moreover, ammonia is said to be carcinogenic. Exposure to harmful levels calls for urine and blood tests. The limit set for exposure to ammonia at open air workplaces is 25 ppm, this being limited to 8 hours of exposure. With regards to short term exposure, this being about 15 minutes, the permissible concentration of ammonia is up to 35 ppm. There are no known risks it has on pregnancy neither does it result in birth defects. Read More