When faced with a drying problem at a plant, the first thing I ask for is a drying curve. Often, people will look at me as if I have three heads —sometimes grumbling: “Why do we need that? We know that drying the material completely takes twenty minutes.” A drying curve can serve several valuable functions; determining the time to reach a certain solvent level seldom is the main one. Unfortunately, one curve doesn’t tell the whole story. The number of curves needed depends on the type of problem.
Drying curves can be expensive to obtain. So, I try to balance my request against the value of the information gathered. There’s no reason to study something unless you know what you’ll do with the results. This underscores the crucial need to identify beforehand the specific data to collect.
The most common drying curve obtained is for oven batch operation. A better option is to use a thin bed of solids to capture something close to single particle drying — this allows for the simulation of many types of dryers. An alternative is to estimate from the oven drying curve the critical and equilibrium solvent content and then use a fluid-bed or rotary device to gather more-precise data. Product stability is best evaluated via differential scanning calorimetry (DSC) to identify any phase changes. Often, this can be conducted with a differential thermal analysis to obtain a rough drying curve. The DSC pinpoints heat flux that may impact the drying process and overall heat requirements.
A batch drying curve can serve for estimating performance in both batch and continuous dryers. For high-value products, consider a small-scale continuous drying test. You can do one by visiting a dryer manufacturer or renting a pilot unit — but it isn’t an appropriate method to generate a drying curve.
You require basic physical properties such as heat capacity, particle size, density and flammability of the solid and liquid. It helps to know any temperature limitation (continuous, short-term), sticky point and minimum fluidization velocity for the material because these determine the potential for product damage and clumping. One factor often overlooked is the ambient environment around the dryer. Dryers need to get rid of the solvent; high ambient humidity or not enough gas flow in the dryer can hinder this. If the gas leaving a dryer is saturated with solvent, you can guarantee that drying is limited.
A drying curve test should include at least three constant inlet temperatures that are below any temperature limitation — except perhaps for products that require a phase change or dehydration. In each test, the inlet conditions (temperature, flow, humidity and pressure) should remain constant.
You can generate the drying curve in three ways: 1) remove a sample as drying proceeds; 2) measure the loss in weight; or 3) calculate the remaining solvent from the exit gas humidity. The latter approach doesn’t disturb the bed of solids and is much more accurate, especially in the later stages of drying. You must adjust loss-in-weight data for buoyancy effects and composition of the loss in weight. Removing a sample for analysis reduces the dry mass of solids under study and makes correcting the overall mass balance difficult. Also, disturbing the bed can mix the solids and promote drying, resulting in an overly optimistic drying rate.
Along with the solvent content as a function of time, you need the temperature of the bed. This may require multiple measurements due to non-uniformity of the gas flow over or through the solids.
No drying curve would be complete without an understanding of the equilibrium moisture content. Determining this can be very difficult because some materials take years to reach equilibrium. A good approach is to maintain a constant humidity on the dried sample for an extended period of time. High humidity is most important because it has the greatest effect on final solvent content of the solids.
From these curves, you can determine the critical solvent content (there may be two or more) and local drying rate along with any limitations due to equilibrium effects. These data are worth having.
TOM BLACKWOOD is a Chemical Processing Contributing Editor. You can email him at TBlackwood@putman.net