Comparison of Methods for Assessing Colloidal Stability of Beer

Abstract

Numerous methods for assessing colloidal stability of beer have been developed. Some of these, such as forcing tests, are mainly used to ensure that beer has been adequately chillproofed; unfortunately, the results are often only available after the beer has left the brewery. The Chapon alcohol cooling test is much faster but requires special equipment and is ill-suited for use as a routine procedure. Other tests, such as the saturated ammonium sulfate precipitation limit (SASPL) and the sensitive (or haze-active) protein test (based on haze induction with tannic acid, followed by turbidimetry) focus on the protein side of colloidal stability and are mainly used in research studies. It was of interest to compare results of these tests on beer treated with different amounts of silicas of different types and particle sizes. Unchillproofed lager beer of a single brand from a single brewery was treated with a number of different silicas, each at three different addition rates (150, 300, and 1,000 mg/L). Each of the resulting samples was subjected to a battery of analytical methods that are often used to evaluate colloidal stability: SASPL, an elevated temperature forcing test, the Chapon alcohol cooling test, and the haze-active (HA) protein test. The HA protein and forcing tests were slightly more precise (lower coefficients of variation) than were the SASPL and Chapon tests. The results obtained with the different methods showed similar, but not identical, patterns with the different silicas; with each method, the three treatment levels with each silica were readily distinguished from each other and from an untreated control. The silicas of different sizes produced generally similar results. The largest particles had slightly lower SASPL and slightly higher HA protein values (both indicating less stability), but in the forcing test, the differences were negligible, with no apparent pattern. The results of the forcing, alcohol-cooling, and HA protein tests showed comparable patterns within the sample set and essentially linear relationships with each other. The SASPL test, on the other hand, showed differences between the samples, but in a different pattern than the other assays. Its relationships to the other methods were distinctly curved. Since the forcing test is the closest to normal practice, it should be the most relevant result. The alcohol-cooling test and the HA protein results were in good agreement with the forcing test, demonstrating their utility in predicting colloidal stability, at least with a single beer brand. As a result, either test should be useful for making comparisons of silica efficacy. The results call into question the utility of the SASPL test for assessing beer haze potential. The short analysis time needed for the HA protein test would permit product testing prior to packaging.