I have covered this subject under different postings in the past [March and April 2007]. However, this is an up to date review with some new information regarding the nature of scaling in black liquor evaporators [1].

Sodium salt scaling takes place in black liquor [BL] evaporators, smelt dissolving tanks and green liquor lines. Scaling occurs when crystals nucleate from solution on pipe walls or heat transfer surfaces. In a solution containing no precipitated salt, nucleation occurs when the solution become supersaturated to the point where the metastable limit for the system is exceeded. The level of supersaturation depends on the solution, the process, the number of hetero-nuclei, the heating rate, the residence time and, sometimes, the shear rate [1].

A salt with normal solubility becomes increasingly soluble when the temperature is raised. A salt with inverse solubility shows the reverse trend. Normal solubility salts typically precipitate-out in cooler regions adjacent to heat exchangers and pipes, and on vessel surfaces colder than the bulk temperature. Inverse solubility salts, however, precipitate on hotter surfaces. Salts can grow by agglomeration of individual crystals or the growth of existing crystals.

Sodium salts scaling in BL

These type of salts precipitate out of BL around 50% solids. The occurrence and degree of fouling depend on the crystal type and the evaporator type. Pilot tests have shown that dicarbonate nucleation, and not burkeite, corresponds to the onset of scaling. The mole ratio of CO₃/ (CO₃₊SO₄) determines the type of the precipitated salt. The salts can be burkeite (1.6-2.6 Na₂SO₄.Na₂CO₃), dicarbonate (Na₂SO₄.1.6-3.0 Na₂CO₃) or sodium carbonate sulphate (1.5 Na₂CO₃.Na₂SO₄) [1].

A rapid method for calculating the solubility limit for the first Na salt to precipitate was developed by Grace [2]. He demonstrated that the concentration of Na, CO₃ and SO₄, and the temperature determine the solubility limit in BL. Grace’s work, combined with a graph showing the ratio of Na₂CO₃/( Na₂CO₃ + Na₂SO₄) in crystal versus the same ratio in solution, can provide a good starting point to determine the type of the precipitated salt and its solubility limit. Although Grace’s work does not differentiate between burkeite and carbonate rich salts, solubility work with dicarbonate in water has shown that it is 3-5% more soluble. So, if dicarbonate precipitates first, the solubility limit will be several % higher than the value obtained by Grace’s method. It has been shown that dicarbonate leads to more significant scaling in falling film evaporators than burkeite, when there is no CaCO₃ scale present. CaCO₃ scales act as a nucleation site for berkeite and may do the same for other Na salts.

The best way to minimize scaling is to have a steady operation and to decrease dead load by increasing reduction and causticizing efficiencies. Operational controls for minimizing scaling include [1]:

· Maintaining a stable crystal population in evaporator bodies which normally operate with crystal present, and avoiding nucleation. To achieve this, the evaporator should be operated at total dry solids well above, or well below (about 2-3%), the solubility limit.

· Sending boilout liquor to a separate tank and returning it into the weak BL as slowly as possible, to prevent a spike in salt concentration.

· Manipulating the ratio of carbonate to sulphate to promote burkeite formation rather than dicarbonate. This can be done by raising the mill sulfidity (within a limit) or by adding ESP dust to BL below the salt system solubility limit. ESP dust addition to the high solids concentrator may be helpful in specific cases.

References:

1. DeMartini, N. and Frederick Jr., W.J., “Review of Sodium Salt Scaling in the Liquid Streams of the Chemical Recovery Cycle of Kraft Pulp Mills”, IPST, Georgia Institute of Technology, Atlanta, GA, USA, Tappi 2008 Engineering, Pulping and Environmental Conf., Aug. 24-27, 2008, Portland, Oregon, USA

2. Grace, T.M. Solubility limits in black liquors. AIChE Symposium Series, 1976. 72(157): p. 73-82.