DaaB wrote:OK I get the gist of that but bearing in mind maths isn't a strong point of mine, how does the formula in your books weigh up, used in conjunction with Tinseths table on boil time vs gravity it certainly seems to have been working for me and is easily adapted for different gravity beers and boil times.
Is it over simplified or could we be endanger of trying to make an exacting science out of something that is still really an educated guestimate? Is it a case of the most accurate method being to stick to a forumla that works with an individual brewing system with the utilisation adjusted based on experience?
If, by the formula in my books, you mean: Weight = EBU*10*Volume/(Alpha*Utilisation)
No. That doesn't change. That's industry standard, universal and absolute. It's the Utilisation bit that is the thorn in the side. I do prefer a fixed figure for utilisation, such as 25%, because it is easy to understand and easy for people to adjust up or down to suit their system. It is more accurate, in my view, than the many fancy formula that abound.
Unfortunately, a fixed utilisation figure doesn't adjust for the alpha contributed by late hops. In the past, I did not consider this to be particularly significant when late hopping with small quantities of noble (low alpha) hops for short periods of time. The advantages of simplicity outweighed the complications of factoring in these small alpha contributions. Anyway, the effect would be swamped by the inherent inaccuracies of the whole hop issue. But now that we have "heretics" that are late hopping with large quantities of high-alpha hops, and hopping at half-a-dozen different times during the boil, these contributions are significant. So I am looking into an easy-to-adjust method of compensating for at least the boil time if not also the gravity factor.
The problem with the commonly-used utilisation formula (Tinseth, Rager, Gareth etc) is that none of them work particularly well, as the article on basic brewing radio demonstrated. There can be 600:1 differences between different methods - clearly that's no good. It has to be said though that the BBR test itself was awful, and it would be hard to draw any real conclusions from the results.
My major complaint with these various formulae, apart from their inaccuracy, is that they are a one-size-fits-all solution, which clearly hop utilisation is not and cannot be. The maths is also too convoluted and, in most cases, just unrealistic cwap from the mathematical point of view. Some have half-a-dozen empirical "frig-factors" in a vain attempt to curve-match poor and insufficient primary information. The end user has no chance of being able to adjust the formula to suit his system in the light of such bodges.
The Tinseth formula most closely matches the theory, in my view, at least the second part does (Time Factor); he just goes a strange way about achieving it. The first part (Gravity Factor) is "sort of" all right, but not good enough. There is clearly a problem with it because if it is plotted out there is an inexplicable "droop" in the middle of the result. It again has too many empirical frig-factors to be adjustable. Every number should be there for a good and valid reason - not just an empirical tweak that makes the numbers look right. From what I can work out I think that Tinseth consistently underestimates utilisation. I think the inaccuracies in Tinseth stem from the Gravity adjustment part. Nevertheless, despite being the best of the bunch, Tinseth did not fare particularly well in the basic brewing radio tests, but nor did any of the others for that matter.
Another tendency is for home brewers (and commercial brewers) to increase the hop rate with gravity on a gravity / EBU ratio basis. This negates, or at least interferes with, the reason for the gravity factor adjustment in these formulae. It is exactly what commercial brewers have been doing for centuries, long before EBU were measured.
It is not really possible for these formula to be particularly accurate; there are far too many variables involved for them to come even close, but they are better than nothing.
So in answer to your original question, I guess that my fixed-figure utilisation is perhaps an over simplification because of late hopping issues. Most of the other solutions are well over the top, because they are trying to do something that they can not really achieve, and some are going about it the wrong way anyway. As you said, people are trying to make an exact science out of a guesstimate. There is almost certainly a happy medium somewhere.
So your approach is the best. Stick with something you know and trust to experience.
.
)