The pilot plant distillation column can not only be used for quantitative performance measurements but can also be made to display very effectively some of the limits of operation of a distillation column. The equipment (see Figure 1) consists essentially of an 8- plate, 80mm, bubble-cap column, thermosyphon reboiler, and two 100 litre vessels, one of which serves as a feed tank while the other collects and combines the top and bottom products during a run. The roles of the two vessels are reversed in the next run. Compositions are determined from the specific gravity (see Table 1), obtained at a measured temperature with a hydrometer. Steam tables will be needed in the laboratory! AIMS 1. Making material and energy balances to check the data. 2. Study of the effect of reflux ratio, feed rate, or boil-up rate on performance. 3. Observation of some of the limits on liquid and vapour flow for satisfactory operation. START-UP PROCEDURE 1. Check that the steam valve by the control panel is shut. Switch on the mains electricity supply. Study the operation of the reflux divider with one person watching the clock on the control panel while the other calls out the solenoid-controlled swings of the glass tube from the platform. Observe that the reflux ratio is the ratio of two times and work out how you can determine it from the settings of the blue hands on the clock. When adjusting the ratio, do not try to force either blue hand of the clock past the red hand; let the red hand move out of the way first. 2. Set the reflux controller to total reflux by switching off the timer when in the total reflux position. Check this from the platform. Set the four receiver manifold valves so that one receiver will function as a feed tank and the other as a collecting vessel for the top and bottom products. Pump enough feed into the column to cover the reboiler coils and up to about 10 cm below the level of the overflow into the still, then switch off the pump. 3. A solenoid valve protects the condenser coils from thermal shock by preventing supply of steam unless cooling water is flowing. Open the water supply valve to the condenser. If no flow is obtained check that all valves in the condenser line are open and the by-pass valve is closed. Adjust the flow to between about 6 and 8 on the rotameter scale. The minimum water flow to prevent the steam supply cutting out is about 5 on the rotameter scale. On the other hand, an excessively large flow will reduce the temperature difference between the water inlet and outlet temperatures (why?) and so give poor precision of measurement of the rate of heat removal by the cooling water. 4. Put the solenoid switch controlling the overhead steam supply to the on position. Crack open the main steam supply valve on the overhead line (i.e. open just a crack), then slowly open it fully and turn back half a turn to prevent risk of jamming (N.B., this sequence is standard procedure for safe operation of any steam valve). 5. Crack open the steam supply valve by the panel and turn it up slowly until the gauge reads about 5 psi. This is to avoid breaking the glass steam coils through thermal shock. Keep UL06/2 the pressure at about 5 psi for 5 minutes while checking that the steam coils start to function properly and observing the progression in the still from natural convection, through sub-cooled boiling, to nucleate boiling. Then slowly open the steam valve fully and close half a turn. Allow the column to heat up at total reflux and observe the upward progress of activity from plate to plate. As the column acquires a liquid hold-up the level in the still will drop well below the level in the bottom product overflow line (why?) but should remain above the level in the coils. 6. Observe the construction and operation of the plates. Each downcomer is sealed by terminating under the liquid on the plate below. This is to prevent rising vapour going up the downcomer and by-passing the plate from which the downcomer descends. A small 'bucket' performs the same function below the lowest plate. The corresponding bucket at the feed plate is missing and you should inspect the activity on the plate above to judge whether it is likely to be as efficient as other plates. EXPERIMENTAL METHOD 1. Steam coils. Look at the steam coils frequently throughout the day to ensure that they are covered with liquid at all times to avoid the risk of explosive breakage of the coils by thermal shock. Discuss with a laboratory supervisor what combination of conditions you would expect to create the greatest risk of uncovering the coils. 2. Heat loss rate to atmosphere. The last part of the column to respond to heat from the reboiler is the top of the column. When constancy of temperature here shows that the whole column has reached steady-state operation at total reflux with no flows of feed or products, determine the rate of heat loss from the equipment to the surroundings by making appropriate measurements. NOTE 1. In subsequent runs the column top temperature can no longer be used as an indicator that the steady state has been reached, since the column is already hot and the azeotrope temperature is not affected by operating conditions other than pressure. 3. Setting a normal operating condition. Now set a steam pressure of about 34 psig, a feed rate of about 0.3 l/min, and a reflux ratio of 4 (instead of total reflux). Note that the feed flow is controlled by the stainless steel needle valve above the rotameter; the blackhandled nylon valve below is only an on-off valve. You may need to re-adjust the needle valve at intervals to maintain a fairly constant feed rate. Start the stirrers in both tanks. 4. Feed and hydrometers. The equipment should initially contain 60-120 litre of about 10% w/w ethanol in water. A kit of two hydrometers, two measuring cylinders and a copper cooling cool is supplied. Sample the feed and determine its density after cooling it to within a very few degrees of the calibration temperature marked on the hydrometer (why?). Since the composition is extremely sensitive to density, read the density with the greatest possible precision. When you take your first hydrometer measurement determine, from the density-composition table posted on the equipment (and Table 1 in this document), the precision of measurement of composition. What is the best precision you can achieve with the hydrometer? If the feed contains less than about 9% by weight of alcohol add more 13% alcohol to raise the feed tank composition to 10-11%. Then intermix the contents of the two tanks so that you have the same feed composition in all runs. Return all samples and liquid run-off from the column to the receiver tank, do not discard.UL06/3 5. Steady-state measurements. Wait until a new steady state has been reached. How can you tell this, bearing in mind Note 1? Then (only then), record temperatures and steam pressure, and all the flow rates that you need (which ones?) to be able to check the data as far as possible by mass and heat balances. Take samples of feed and top and bottom products after running off any unequilibrated contents of the sample lines (not to be confused with the contents of the graduated vessel in the top product line. If you empty this vessel below the zero on its scale it will take a long time to fill up and allow you to take measurements again!). Return all samples and run-off liquid to the receiver tank. Note the following important points. (a) Flow measurement. Some of the flows you may want to measure are intermittent or unsteady even when temperature readings indicate that the steady state has been achieved. In this case you need a time-averaged flow rate. Collect liquid over three or four successive but equal intervals of time. If the flow variation is too large, double the time interval. Repeat until you find an interval long enough to give a reliable flow rate. (b) Zero error of thermocouples. Is there any evidence of a zero error in the thermocouple temperatures? If so, it is likely to be in the electronic reference cold junction which is common to all the thermocouples. Devise a way of determing the zero error correction near the low end of the temperature range in the equipment by measuring one of the thermocouple temperatures by an independent method. At the high end of the temperature range in the equipment you can determine the error more approximately by comparing the temperature registered in the reboiler with the temperature read from steam tables at atmospheric pressure. This assumes, first, that the reboiler pressure is atmospheric and, second, that the contents of the reboiler are close enough in composition to pure water not to affect the boiling point significantly. You should be able to estimate the effect of the second assumption from your composition data. Once you have determined the zero error, use it to correct all temperature registered by the thermocouples. (c) Make sure that you collect all the data you need for your calculations. Check this by completing the calculations on your first run (reflux ratio 4) as early as possible while you are doing the next run and still have time to correct any oversights. 6. Subsequent runs. Determine the effect of one or more operating variables. Do not attempt more runs than you are going to have time to calculate up. Three or four runs are suggested (provided you ensure that equilibrium is achieved in each) and are probably best devoted to varying the reflux ratio. A suitable range of reflux ratios to attempt is 1-8. Within this range the reflux ratios should be spaced non-linearly, e.g. in geometrical progression, for good experimental design, since changing the reflux ratio from 1 to 2 creates a much larger percentage change in the column's internal flow rates than changing from 7 to 8. The steam pressure may need some adjustment because the heat demand changes as the flow rate of liquid to the reboiler changes with reflux ratio. 7. Limits of operating conditions. A distillation column can function only within a limited range of ratios of flow rates of feed, falling liquid and rising vapour. A column does not work under operating conditions which produce flooding, weeping, etc., or lack of a top or bottom product.UL06/4 First, try flooding the column by raising the steam pressure and feed rate to maximum at a high reflux ratio. DO NOT let flooding build up to a greater depth than two full plates or you will expose the steam coils: turn off both the steam supply and feed completely to stop flooding. Why did flooding begin on the lowest plate? What happened to activity on higher plates at the same time, and why? When the excess liquid hold-up has subsided back into the reboiler, try setting a steam pressure of 20-25 psi and restoring the previous feed rate. When a steady state is again reached, which product stream have you lost and why? What would happen to the product flows if the steam pressure were high enough for the rate of production of vapour in the reboiler to exceed the liquid flow rate in the stripping section, assuming the reflux ratio and feed rate were low enough to avoid flooding? SHUT-DOWN PROCEDURE 1. Close main steam valve (on overhead line). 2. Put solenoid switch to off position. 3. Close steam valve on equipment. 4. When ebullition has stopped, close water valve on equipment. 5. Switch off mains electricity supply. THEORY 1. Mass Balance. A question arises: is the flow rate of bottom product (not instrumented) needed to test the mass balance? Conservation of mass of total material and ethanol gives two equations, in the notation of Reference 1: F D W Fx Dx Wx f d w = + = + ( ) ( ) 1 2 Eliminating the unknown flow rate of bottom product W between these two equations gives the single equation Fx Dx F D x f = d + - w ( ) (3) All the parameters in this equation can be measured. The flow rate of bottom product is therefore not needed to permit the mass balance to be tested. You can calculate the values of the two sides of the equation independently and compare any discrepancy with the precision of measurement. This mass balance provides an important test of the validity of the data collected. 2. Simplifying assumptions. The ethanol + water system exhibits substantial positive deviation from ideal solution behaviour (see Figure 2). The heat of mixing and volume of mixing are therefore not zero. They are, however, much smaller than in systems such as hydrogen chloride + water or ammonia + water, and can be treated as zero for most calculational purposes. Treating the volume of mixing as zero allows you to calculate the volumetric flow rate of a mixture as simply the sum of the volumetric flow rates of the two pure components. Treating the heat of mixing as zero satisfies one of the preconditions for the assumption of constant molal overflow used in determining the number of plates in the column, the other precondition being negligible heat loss from the column. When you are UL06/5 determining the number of plates, the vapour-liquid equilibrium diagram (Figure 2) for aqueous ethanol can be copied as needed. 3. Consistent Units. During your calculations you will often need to calculate the flow rate of one component present in a flow of mixture by multiplying the flow rate of mixture by the fraction of that one component. When you do this, the two quantities multiplied together must be in consistent units e.g. a volumetric flow rate must be multiplied by a volume fraction, and a molar flow rate by a mole fraction. To convert volume fractions to mole fractions you will find it easiest to start with a basis of a volume (e.g. 100 ml or litres) of mixture. RESULTS AND DISCUSSION Establish the accuracy and precision of the calculated results. 1. Validation of data (a) It is normal practice with process plant to conduct material and energy balances to check for significant errors in instrumentation, measurement and calculation. Make overall mass and heat balances for operation at one of the finite reflux ratios used, i.e. not total reflux. The mass balance is described above. The heat balance is best made as follows. Consider your calculated values for: · the heat input rate from steam condensing in the reboiler (i.e. given up as latent heat of condensation), · the heat transfer rate to the cooling water, · the net heat transfer rate to the process streams, and · the heat loss rate to the surroundings (determined at total reflux). Consider also your estimates of the precision of measurement of these quantities. Express any discrepancy in the balance as a percentage of the heat input rate in the reboiler. Try to explain the discrepancy if it is significant in comparison with experimental error. Justify your assumption, in this heat balance, that the heat loss rate at any reflux ratio is likely to differ little from that at total reflux. NOTE 2. The above way of conducting a heat balance is preferable to the alternative method of computing the total enthalpy flow into the column from all streams entering (steam + feed +cooling water) and total enthalpy flow out (steam condensate + distillate + bottoms + cooling water + heat loss). With the recommended method it is easy to assess the significance of any discrepancy between the two sides of the balance by comparing the discrepancy with the heat transfer rate. No such comparison is possible with the alternative method since the total enthalpy flows in and out depend on the arbitrary reference temperatures used to define the enthalpy scales (the discrepancy itself should, of course, be the same by either method.) (b) Express the heat loss rate as a proportion of the heat input rate in the reboiler. Say what implication this value has for the assumption of constant molal overflow and hence for the plate numbers you are determining by the McCabe-Thiele method. 2. Experimental variables By means of graphs, show the effect of the reflux ratio (or other experimental variable(s) studied) onUL06/6 (i) the product purity (ii) the column efficiency (ratio of ideal stages required to actual plates present). Interpret effect.