The re-discovery of first wort hopping

F. Preis, LGA Nuremberg, and W. Mitter, Simon H. Steiner, Hopfen GmbH, Au/Hallertau.

The re-discovery of first wort hopping came about by accident. In a particular brewery operation, part of the hops were accidentally added to the first wort. After this was noticed, the brew in question was run separately. When the storage cellar beer and later ton the finished beer was tasted, a hop aroma different to that of the normally hopped beers produced by this brewery was noted. his observation aroused a certain interest. A literature search showed that first wort hopping is not a new discovery as it was quite common at the turn of the century. But the aim at that time was not to improve the aroma, but to achieve a better bitterness utilisation. The first step towards first wort hopping consisted in a hop screen through which the spargings were passed.1 But as the quality of these beers was deficient in terms of bitterness, an attempt was made to dose a third of total hop addition to the first wort.2,7 Schönfeld recognised that there was a connection between the improved bitterness utilisation and the higher pH of the first wort when isomerisation of the alpha-acids is better.3 In the forties,Kolbach andWilharm carried out tests in connection with mash hopping and recognised high bitterness losses.4 Schur andPfenniger confirmed these losses in tests in connection with mash hopping but their beers had a fine and clear smell and taste of hop aroma.5 When reworking these tests,Gresser neither found a hop aroma nor higher hop oil contents.6 Lense andReitmeier report of a process where the whole hops are placed onto the first wort without stirring in order to avoid cooling.1 In his "Brewers' Manual",De Clerk describes a process which is also very similar to first wort hopping. The hops are immersed in 50°C warm water before being introduced into the copper. This was to remove unpleasant substances and smells.7,9 In his thesis, Gresser confirmed the loss of aroma components in the volatile range.6 Panglisch did this pretreatment with hop extract and used water of a temperature of 70°C. He found a stronger hop aroma in the beer.8 Commercial tests in connection with first wort hopping Test set-up Test brews had already been run quite successfully in various breweries. It was therefore decided that the next tests to be run in two more breweries would include extensive analytical measurements of the whole procedure. The tests were run with Pils beers as hops has a major influence in production of these beers. In order to obtain comparable results, the breweries selected had to fulfill the following criteria: • identical pitching rates and number of cycles; • identical malt from one and the same batch; • comparable duration of wort boiling; • approximately the same composition of brewing water.

First wort hopping was carried out when the copper bottom was covered with liquid. The hop products, in this instance pellets types 45, were not stirred in. Tables 1 and 2 list the technical equipment in the two breweries and the process sequence followed during first wort hopping. In brewery A (Table 1), what was normally the last addition comprising Tettnanger and Saazer type 45 was added to the first wort. This represents a quantity of 34% of total hop addition. In brewery B (Table 2), even 53% of the total hop addition was added to the first wort in the form of Tettnanger type 45. In both breweries, subsequent aroma hop addition was dispensed with. It is worth mentioning that this type of hopping did not represent any problem in the individual brewery operations.

Table 1: Procedure in brewery A Brewhouse equipment: conventional automated, copper directly fired. Wort boiling time: 90 min without pressure Biological acidification: none Cold store: cold wort filtration Fermenting room: flat conical tanks Fermentation conditions: cold Storage cellar: horizontal tanks Filtration: kieselguhr filter/sheet filter Total Hop addition: 13 g of alpha-acid/hl of cast wort. 34% of this (corresponding to the two last additions originally used) into the first wort in the form of Tettnanger type 45 and Saazer type 45 in the test brew. Table 2: Procedure in brewery B Brewhouse equipment: conventional automated, internal boiler Wort boiling time: 80 min without pressure Biological acidification: none Cold store: flotation Fermenting room: flat conical tanks Fermentation conditions: cold Storage cellar: horizontal tanks Filtration: kieselguhr filter/sheet filter Total Hop addition: 13 g of alpha-acid/hl of cast wort. 52% of this (corresponding to the two last additions originally used) into the first wort in the form of Tettnanger type 45 in the test brew.

Beer analyses Table 3 and 4 show the beer analyses of both breweries. It can be seen that the reference and test beer have approximately the same composition in both breweries. The finished beers of brewery A had slightly different original gravities. The test brew had an insignificantly higher final attenuation of the finished beer. It is very obvious that the beer produced by first wort hopping had an improved foam index particularly in brewery A. It cannot yet be said whether or not this is due to first wort hopping, the differences are too small. Further tests would have to be run. The slightly higher iso-humulones content in the beers produced by first wort hopping would be a possible explanation. It is known that the iso-α-acids contribute to a better foam stability. The tannins might also play a certain role, these are leached more intensively by first wort hopping and might also have a positive influence on the beer foam.

Table 3 Beer analyses brewery A   Reference brew Test brew Original gravity (wt. %) (vol. %) 11. 13 11. 62 11. 50 12. 01 Alcohol (wt. %) (vol. %) 3. 76 4. 80 3. 99 5. 09 Real extract (wt. %) 3. 81 3. 74 Apparent extract (wt. %) 2. 10 1. 93 Spec. gravity at 20°C 1. 00820 1. 00750  Final attenuation of the finished beer: Apparent (%) 81. 10 83. 20 Real (%) 65. 80 67. 50 pH 4. 56 4. 53 Carbon dioxide content (wt. %) 0. 57 0. 60 Foam acc. to Ross & Clark 113 123 Table 4 Beer analyses brewery B   Reference brew Test brew Original gravity (wt. %) (vol. %) 11. 49 12. 00 11. 43 11. 93 Alcohol (wt. %) (vol. %) 3. 72 4. 75 3. 67 4. 69 Real extract (wt. %) 4. 26 4. 29 Apparent extract (wt. %) 2. 59 2. 63 Spec. gravity at 20°C 1. 01009 1. 01027  Final attenuation of the finished beer: Apparent (%) 77. 5 77. 0 Real (%) 62. 9 62. 5 pH 4. 58 4. 56 Carbon dioxide content (wt. %) 0. 55 0. 56 Foam acc. to Ross & Clark 121 124

Tasting results The tasting results are particularly interesting (Tables 5 and 6). The tastings consisted of the triangular test. Each tester gets three samples, two of which are identical. The beers must be categorised correctly and the overall impression must be subsequently assessed. This method makes it possible to objectively assess the beers. In brewery A, eleven of the twelve trained tasters found the correct category. Three of these eleven tasters preferred the reference sample and eight favoured the sample produced by first wort hopping. This result corresponds to a significance of 99 percent. The result was even more clear-cut in brewery B. Twelve of the thirteen tasters arrived at the correct categorisation, eleven of the twelve tasters preferred the sample produced by first wort hopping. This result corresponds to a significance of 99. 9 percent. The tasters gave the following main reasons for their preference for the beers produced by first wort hopping: • a fine, unobstrusive hop aroma; • a more harmonic beer; • a more uniform bitterness No results are yet available regarding the taste stability of the two beers. Studies are being undertaken to clarify this aspect; these studies will be the subject of a separate article.

Table 5 Results of tastings brewery A Triangular taste testing Reference sample = O-P First wort hopped sample = VW-P Taster Correct categorisation Preference 1 yes O-P 2 no - 3 yes VW-P 4 yes VW-P 5 yes VW-P 6 yes VW-P 7 yes VW-P 8 yes O-P 9 yes O-P 10 yes VW-P 11 yes VW-P 12 yes VW-P Wrong categorisation 1 Correct categorisation 11 Of the 11 correct categorisations: Preference O-P 3 Preference VW-P 8 2-star-significance (99% reliability) Table 6 Results of tastings brewery B Triangular taste testing Reference sample = O-P First wort hopped sample = VW-P Taster Correct categorisation Preference 1 yes VW-P 2 yes VW-P 3 yes VW-P 4 yes VW-P 5 yes VW-P 6 yes O-P 7 yes VW-P 8 yes VW-P 9 yes VW-P 10 yes VW-P 11 no - 12 yes VW-P 13 yes VW-P Wrong categorisation 1 Correct categorisation 12 Of the 12 correct categorisations: Preference O-P 1 Preference VW-P 11 3-star-significance (99.9% reliability)

Analyses of bitter substances Table 7 shows bitter substance analyses for worts and beers of breweries A and B. It is obvious that worts of the reference brews contain considerably more non-isomerised alpha-acids than worts from first wort hopping. This is even more apparent in brewery B which brewed with a significantly higher addition to first wort. An analogously very much higher iso-alpha-acid content in the worts was apparent, this was still relatively clearly identifiable in the beers. The bitterness units according to EBC also reflect this picture in the beer. Therefore it is the more surprising that these beers with increased bitterness were assessed as being finer and more pleasant in tastings. It is usually a good sign when the bitterness of a beer is sensorially less pronounced than should be on a purely analytical basis.

Table 7 Analyses of bitter substances in worts and beers Hop addition: Brewery A: 13. 0 g of α-acid/hl of cast wort, 34% of this to first wort. Brewery B: 12. 2 g of α-acid/hl of cast wort, 52% of this to first wort Analyses Brewery A Iso-α-acid α-acid EBC BU Brewery B Iso-α-acid α-acid EBC BU Worts Reference brew 47. 6 15. 2 - 32. 6 28. 4 - First Wort Hopping 55. 0 10. 6 - 44. 8 12. 3 - Beer Reference brew 40. 9 2. 9 37. 9 27. 4 3. 3 27. 2 First Wort Hopping 42. 7 2. 4 39. 6 35. 1 3. 3 32. 8

Gas chromatographic tests and results The aroma spectrum of the worts and beers was determined using gas chromatographic analysis. The test methods used have already been described.12 Figures 1 and 2 are representative of chromatograms of the reference brew and the beer from brewery B brewed with first wort hopping. The peaks of linalool, terpineol, gerianiol and humulene epoxide are identified, substances which the literature partly connects repeatedly with hop aroma.6 11 For further assessment, hitherto unknown peaks will also be reproduced in order to complete the overview.

Figure 1 Beer from reference brew Figure 2 Beer from first wort hopped brew

Tables 8 and 9 show the concentrations of aroma substances in mg/l for the utilisable aroma substances. These tables also list hexanol, ethyl octanoate, and phenylethyl acetate, substances which are only formed during fermentation. As in the beers produced by different hopping processes, the three last named substances showed very good agreement, it is obvious that as far as the other parameters are concerned the brews were very uniform.

Table 8: Concentrations of aroma substances µg/l brewery B   Linalool Terpineol Geraniol Humulene epoxide Hexanol Ethyl octanoate Phenylethyl acetate Reference wort 24. 7 6. 3 12. 0 25. 5 - - - First wort hopped wort 1. 2 - 11. 0 6. 9 - - - Reference beer 34. 1 5. 3 14. 6 10. 8 15. 2 110 584 First wort hopped beer 6. 4 - 13. 7 9. 8 15. 8 118 606

Table 9: Concentrations of aroma substances µg/l brewery A   Linalool Terpineol Geraniol Humulene epoxide Hexanol Ethyl octanoate Phenylethyl acetate Reference wort 12. 3 ? 11. 9 11. 2 - - - First wort hopped wort 7. 2 ? 12. 1 19. 0 - - - Reference beer 29. 0 6. 9 18. 8 32. 7 18. 2 158 725 First wort hopped beer 8. 1 4. 5 10. 7 19. 6 11. 7 173 749

But differences between reference worts and beers with first wort hopping are very marked. The reference brew especially in brewery B i. e. , the brewery working with a very high first wort hopping rate hardly contains any linalool while a considerable quantity of this substance is still present in the reference brew. The situation is quite similar as far as terpineol is concerned. This substance is present in the reference brew but has completely disappeared from the brew produced with first wort hopping. Approximately the same applies for the two substances in beer: In both cases, linalool has slightly higher values than in the wort but the difference between reference beer and first wort beer is again very pronounced. This also becomes apparent from Fig. 3 which shows a section of the two chromatograms overlaid for the two beers from brewery B. The higher peak represents the internal standard, the peak to the right is linalool. The small peak ahead of the standard is also remarkable. Again, it is more pronounced in the case of the reference beer (green marking).

Figure 3 Brewery B, linalool peak in reference and first wort hoppped brew green marking - reference brew red marking = first wort hopping Figure 4 Brewery B, humulone epoxide peak in reference and first wort hopped brew green marking - reference brew red marking = first wort hopping

The same trends were apparent in brewery A but naturally the differences were not so pronounced. As in brewery B, the linalool content rose from wort to finished beer in brewery A. In his study, Gresser 6 also observed an increase in the linalool content during fermentation. The difference between reference brew and first wort hopping was very much more pronounced in the beer than in the wort. In brewery B, the geraniol content is practically identical in both brews in the wort as well as in the beer. Humulene epoxide was still very high in the reference brew compared to the brew produced by first wort hopping. This situation has equalised in the finished beer where hardly any difference can be noted, as can be seen in Figure 4. This chromatogram shows a peak at 45. 4 min which can be clearly identified only in the reference brew. The same situation arises in brewery A. There is practically no difference in terms of geraniol. Humulene epoxide shows a slightly different picture. While a lesser quantity is present in the reference wort, the humulene epoxide content is higher in the reference beer compared to the first wort beer. When looking at the section of the chromatogram in Figure 5 in which once again the beer from the reference brew and from the first wort hopped brew of brewery A are overlaid, it is apparent that also in this section (oxygen fraction) hop aroma substances are definitively higher in the reference brew. These substances are unknown to us but they might also have an influence on hop aroma and on taste.

Figure 5 Brewery A, oxygen fraction of reference and first wort hopped brew green marking - reference brew red marking = first wort hopping Figure 6 Brewery B, oxygen fraction of reference and first wort hopped brew green marking - reference brew red marking = first wort hopping

It has been mentioned that a substantially larger hop quantity was added to the first wort in Brewery A. It is interesting to note that the differences in terms of oxidised aroma substances are not as large as in brewery B (Figure 6). Although certain tendencies towards higher peaks are observed in the normal brew, the differences are not very pronounced. In conventional hopping, a number of smaller peaks also arise, these do not occur at all or just very weakly in the test brew. But it must be recognised that this group of higher boiling aroma substances is generally clearly lower in brewery B than in brewery A. The fact that an additional internal boiler is used possibly contributes to this development. It should be borne in mind that the analyses of the twin brews yielded exactly the same results, this also reflects the fact that the parameters of the test brews in the brewery were exactly adhered to. When comparing the two operations, it becomes obvious that the normally hopped beers always contain more or a higher content of hop aroma substances. This finding is quite surprising as the first wort beers were assessed as having a very fine and rounded hop aroma and a rounded hop taste. In the region of oxidised aroma substances, the oxygen fraction, either new or higher peaks which have already formed int he first wort at lower temperatures would have been expected. The very low linalool value in the test brew compared to the reference brew is also surprising. This component is oftentimes connected with a typical hop aroma. On the other hand, it appears to be quite reasonable that the quantity of linalool in the first wort hopped brew is lower because theis component had been subjected to a longer boiling time. The same applies to other aroma components which are practically subjected to steam distillation during wort boiling and thus are distilled off to a large degree. This has been postulated byAnderegg some time ago.11 When combining these observations with sensorial assessments, one reason for the better ranking of the beer produced by first wort hopping could be that the absence of a number of aroma components or a lower concentration of certain components might have a positive effect on aroma and also on taste. It is very difficult to describe a hop aroma analytically. We are just considering assumptions which must not be carried too far after tests in only five breweries, in only two breweries of which were the tests accompanied by comprehensive analyses. Our own tests run with hop oil additions did not yield any significant findings in terms of analysis and taste.12 Every brewery should in principle carry out its own tests in order to include the influence of its own operational parameters to be able to get a picture of the taste impression. But we recommend that first wort hopping be carried out with at least 30% of the total hop addition, using the later aroma additions. As far as the use of hops is concerned, the alpha-acid quantity should not be reduced even in the case of improved bitterness utilisation. The results of the tastings showed that the bitterness of the beers is regarded as very good and also as very mild. A reduction of the hop quantity added could result in the bitterness being excessively weakened, and the good "hop flavor impression" could be totally lost. First wort hopping might bring good results in combination with biological wort acidification in order to somewhat compensate for the loss of bitter substances and simultaneously produce a high-quality beer. References Lengse, K. , Reitmeier, R. : Katechismus der Brauereipraxis, 14th ed. , 1970. Wiegmann, D. : Allgemeine Brauer u. Hopfenzeitung, 1912/1913, pp. 233, 663, 799. Schönfeld, F. : Handbuch der Brauerei und Mälzerei, vol. I, 1930. Kolbach, P. , Wilharm, G. : Wo. Br. 1943, 66. Schur, F. , Pfenniger, H. : Schweizer Brauerei-Rundschau 85, 65-84, 1974. Gresser, A. : Dissertation Technische Universität München, 1985. De Clerk, J. : Lehrbuch der Brauerei, vol. I, 1950. Panglisch, P. : Dissertation Technische Universität München, 1988. Lüders, H. : Die wissenschaftlichen Grundlagen von Mälzerei und Brauerei, 1949. Esslinger, H. M. : Brauwelt 129, 945, 1989. Anderegg, P. : Schweizer Brauerei-Rundschau 87, 1-64, 1976. Mitter, W. , Kessler, H. , and Biendl, M. : Brauwelt 133, 979-986, 1993.

The authors:F. Preis, LGA Nuremberg, and W. Mitter, Simon H. Steiner, Hopfen GmbH, Au/Hallertau.

This article copyright 1995, Brauwelt International. HTMLized and reprinted by permission, 2002.