Partial condensers may be used with distillation columns in two modes. First, the vapor may be the only overhead product, with all of the condensed liquid returned to the column as reflux. Second, a portion of the condensed liquid may be withdrawn as a top liquid product in addition to the vapor product.
The second mode is usually used when there are essentially non-condensable gases that can be easily excluded from the top liquid product by the single condenser stage. The vapor product becomes a waste stream and the liquid distillate becomes the overhead product. This allows the condenser to operate at a reasonable temperature because the light gases don't need to be condensed.
The Prode Properties DCOL routine can simulate total condensers and partial condensers with no liquid product. It can also simulate a partial condenser with both liquid and vapor products by using a liquid side draw stream on the top, condenser stage. In this mode of operation, you specify the ratio of the side draw rate to the total feed rate. The extra degree of freedom provided by the draw stream is automatically used by this specification of the draw rate.
The normal definition of reflux ratio is the rate of reflux to the overhead products (vapor and liquid). Because DCOL doesn't recognize the condenser stage side draw as an "overhead product" when a partial condenser is being simulated, it's reflux ratio is defined as the reflux over vapor product rate. In my Mathcad column routine, I input the "true" reflux ratio and the routine computes the ratio according to the DCOL definition when a partial condenser with side draw is being simulated. Note: the mc_DCOL routine does not make this change. The change is made in my special routine that I created to do case studies.
Convergence with large liquid distillate product
It may be difficult to obtain convergence when you want the liquid overhead product to be a large fraction of the total overhead. If you have a problem, look at the light key concentration stage profile and see if your feed location is correctly placed. By changing my feed location, I was able to increase the amount of liquid product and achieve convergence. Probably the convergence difficulty is related to the column initialization.
If you are simulating a real column (i.e. not a design), then you can't change the feed location. Maybe manual initialization will allow the program to converge with the large distillate draw stream?