When your basic column controls fail, a quick thinking engineer can save a company alot of money. Weighing the financial options before the incident happens can help you feel secure in you decisions when you make them. We'll have a look at three scenarios that you may not have thought of, but may encounter.
Let's start with a basic, optimizedcolumn design as described in Table 1:
Table 1: Column Design Specifications
|Feed Temperature (0C) |90.0 |
|Feed Pressure (kPa) |19.0 |
|Top Pressure (kPa) |6.5 |
|Bottom Pressure (kPa) |15.8 |
|Reflux Ratio |12.2 |
|R/Rmin |1.18 |
|Feed Stage |9 ||Number of Actual Trays |31 |
|Tray Efficiency |78% |
|Diameter (m) |1.2 |
|Length (m) |22 |
|Flooding |53% |
|Condenser Heat Duty (MJ/h) |-2726 |
|Reboiler Heat Duty (MJ/h) |2679 |
|Acrylic Acid Purity |99.9% |
Figure 1 shows a materialbalance around the column:
Situation: Following a short time during which capacity was increased, there was some difficulty in adjusting the utility flow rates in order to get the column to operate properly. It seems that the column is currently using more cooling water and steam than the current feed flowrate should require to obtain the necessary purity according to thecolumn design. The utilities had to be taken off of automatic control and adjusted manually causing extensive amounts of downgraded material because the purity specification was not being met.
Consequences: Lost revenue due to off specification material.
Cause: During the scale up, the utilities are increased to handle the additional heat duty. When production returned to normal, the steamflow in the reboiler was reduced according to the feed flow, not the original conditions before the scale up (although there should be no difference). However, the design utility flowrates can seldom accurately predict fouling. This means that more utilities may be required than specified by the design specifications. This is the case here.
Solution: After increasing the steam flow to it'sproper set point, refine the operating procedure for capacity changes. The correlation between feed flow and steam may need to be re-examined. Over time, the steam flow may gradually increase to maintain purity. Mineral buildup in the reboiler (fouling) may need to be accounted for in the correlation. For example, the feed flow-utility flow correlations may also need to be time dependent toaccount for the fouling in the reboiler/condenser. Examining the column's history may be helpful here. You may be tempted to say, "Why don't we just use the set points that we used before the scale up?". That will work if you are returning to exactly the same flowrate, but you may be required to change capacity to flowrate that you don't have a reference point for in the immediate past.
Situation: The acrylic acid is stored in holding tanks until shipment to customers. The purity is tested in the lab everyday and continuously online. Currently, one of the acrylic acid tanks is off specification at 99.5% pure. The condenser and reboiler utility streams have been behaving strangely by oscillating up and down in the past two days. Because these oscillation have beensmall and barely outside of set point parameters, no changes were made. The concentration detector that measures the online concentration of the acrylic acid has been operating properly.
Consequences: If shipped as is, this material would have to be downgraded resulting in a serious financial loss.
Cause: All to often, engineers forget about one simple upset in columns. Feed surges can...