The Surprising Science of Dry Hopping – Lessons from Tom Shellhammer

Last week I attended parts of the Ohio Craft Brewer’s conference, held at the Hyatt Regency in downtown Columbus. For me the most interesting aspect of the conference was Tom Shellhammer’s keynote address [1]. Shellhammer, the Norwester Professor of Fermentation Science at Oregon State University, is internationally recognized for his research on hop chemistry. His presentation at OCBC focused entirely on the chemistry of dry hopping (adding hops to the beer during conditioning and/or fermentation). As readers of this blog will know, dry hopping is a brewing technique that figures prominently in many styles, most notably American IPAs of every description. Shellhammer’s talk featured three vignettes, each of which had a clear take home message that I attempt to summarize below. I’ve done my best to keep the chemistry at an accessible level.

tom shellhammer
Dr. Tom Shellhammer during his presentation at the Ohio Craft Brewers Conference.

When more is not always more

The first segment of the talk asked the question, does using larger quantities of dry hops lead to more hop aroma? To answer this question Shellhammer and his colleagues started with base pale ale (4.75% abv) that was very lightly hopped (19 IBU), and dry hopped it with whole cone Cascade hops for 24 hours [2]. The variable they chose to explore was the quantity of dry hops, which ranged from 2 g/L to 16 g/L. Translated into units that would be familiar to metrically challenged American homebrewers, the lowest rate of 2 g/L translates to 1.3 ounces per 5 gallon batch (0.5 lbs/bbl), while the highest rate, 16 g/L, is a whopping 10.7 ounces per 5 gallon batch (4 lbs/bbl). The effects of dry hopping on the finished beer were analyzed using both analytical instrumentation and a trained panel of tasters.

The first conclusion is that above 8 g/L the extraction of hop oils and the aroma contributions from those oils are nearly saturated. To use Shellhammer’s language dry hopping with more than 8 g/L (1.1 oz/gal or 2.1 lbs/bbl) is an inefficient use of raw materials [3]. As a chemistry professor this strikes me as reasonable conclusion. As Tom noted during his talk, the most aromatic components of hop oil have the lowest solubility in water, which limits the amount you can extract. His second finding was more surprising. The oils responsible for citrus aromas saturate earlier, closer to 4 g/L, than the oils responsible for herbal/tea aromas and the compounds that impart bitterness (humulinones and polyphenols). The implications are best summed up in the conclusion of a paper published by Shellhammer and his student Scott LaFontaine [3]:

Adding more hops by static dry-hopping does not simply lead to increased aroma intensity but also changes aroma quality in the finished beer. Dry-hopping rates >8 g/L lead to hop aromas that were more herbal/tea in quality than citrus. To maintain a more balanced hop aroma quality this study suggests using a static dry-hopping rate between 4 and 8 g/L.

How does this compare with dry hopping rates found in practice? I’ve read that Russian River uses approximately 1 lb of dry hops per barrel in Pliny the Elder. While that seemed like a ridiculous amount in the early part of this century, nowadays it doesn’t seem out of the ordinary. Shellhammer’s research would seem to indicate this is still in a range that is useful. Personally, the largest charge of dry hops I’ve used is 4 ounces in a 5 gallon batch, but my buddy Chris Mercerhill once dry hopped with a pound of hops just to see how it would turn out. That experiment, dubbed “The Pounder” also confirmed the notion more is not always better.

dry hop aroma characteristics shellhammer

A figure taken from reference [3] showing how the herbal/tea like aromas (gray circles) continue to increase after the citrus aromas (white circles) saturate.

Does dry hopping lead to bitterness?

Until very recently brewers and brewing scientists attributed hop bitterness almost entirely to isomerized alpha acids that are produced in the boiling wort. The alpha acids found in hops are themselves not particularly bitter and have very little solubility in water, but in the harsh conditions of the boil they are isomerized, which in layman’s terms means that the molecule changes its shape (and polarity) by rearranging its atoms. This change increases both bitterness and solubility [4]. At fermentation temperatures where dry hopping occurs there is not enough heat to isomerize the alpha acids, so the conventional thinking was that dry hopping does not add bitterness. For example, the homebrewing calculator that I use assumes no IBU contribution from dry hopping.

With the recent trend toward intensely fruity IPAs with minimal bitterness, brewers have reduced the hops added to the boil (in some cases to almost nothing), while increasing the dry hopping rates significantly. This unprecedented approach to brewing has led brewers and scientists like Shellhammer to rethink the old assumptions. His research shows that compounds called humulinones and polyphenols can make significant contributions to bitterness in heavily dry-hopped beers [5]. Unlike the isomerized alpha acids these compounds do not need to be boiled to become bitter. Furthermore, because the humulinones are formed when (non-isomerized) alpha acids are oxidized, this effect is more dramatic when using old hops that have partially oxidized. What’s the take home message? Don’t put much stock in the low IBU numbers associated with hazy/juicy NEIPAs and if you do want to minimize bitterness seek out brewers that dry hop with the freshest possible hops.

Hop Creep

The final, and arguably most interesting, segment of the talk explored the phenomenon of hop creep. Never heard of hop creep, don’t feel bad, you would have had plenty of company in the audience at the Ohio Craft Brewer’s Conference. It refers to the process whereby additional fermentation is triggered by dry hopping. Shellhammer described an experiment, which is summarized in the graph below taken from his 2018 article in the Journal of Agricultural and Food Chemistry [6].

hop creep shellhammer

This plot summarizes Shellhammer’s fascinating experiments with hop creep in Coors Banquet Beer. It was taken from reference [6].

In this study he took four samples of Coors Banquet beer. To one beer he added Cascade hop pellets, to another he added yeast, to a third he added both hops and yeast, while the fourth was not modified. The y-axis of the graph above represents the levels of unfermented sugars in the beer (gravity in brewer’s parlance). The sugar level in the control sample shows almost no change with time, holding steady at 3.5° Plato for 40 days. The beer to which yeast was added shows a modest decrease in sugar levels, perhaps not too surprising, after all Coors doesn’t make the world’s driest beer. The beer to which hops are added shows a slight rise in sugar levels that holds fairly constant for the duration of the experiment. It turns out there are low levels of sugar in hops that can explain this rise. The real shocker is the fourth beer, where both hops and yeast are added. The gravity (sugar levels) of that beer drop by 50% to ~1.7° Plato over the 40-day duration of the experiment!

The important take home point is that additional fermentation only occurs when you add both yeast AND hops. What in God’s green earth could explain this behavior? It turns out there are enzymes in hops that can break down the unfermentable long-chain sugars to simple sugars, much like the enzymes in barley that carry out the same function during the mashing stage. This enzymatic activity doesn’t make much difference if there is no yeast present (sugar is sugar), but in the presence of yeast the newly produced simple sugars are subsequently fermented to ethanol and CO2. Hop creep is only an issue in dry hopping, because if the hops are added to the boiling or near-boiling wort the high temperatures will denature (destroy) the hop enzymes.

The implications of hop creep can be significant, especially for hazy NEIPAs which often contain suspended yeast and depend on high levels of dry hopping to pack in the hop aromas:

  • If the beer has already been packaged the additional CO2 produced during fermentation can lead to exploding cans/bottle bombs.
  • A boost in abv will be observed, in the Coors Banquet Beer experiment the abv increased by 1.3%
  • Additional diacetyl (an unwanted off-flavor that tastes of imitation butter) is produced. Tests in Shellhammer’s lab showed diacetyl levels rose from 25 ppb to 200 ppb due to the bump in fermentation triggered by dry hopping.

The more unfermented sugar in the beer the more problematic hop creep becomes, which means you’re more likely to experience an exploding DIPA can than a Brut IPA.

Think this is just an academic exercise? Try googling the words “exploding beer cans” and see what you find (or read Bryan Roth’s article on Good Beer Hunting). After the conference I was speaking with Mark Richards from Land-Grant Brewing and he told me that recently they had to hold off packaging their most recent hazy IPA (Piña Pants) for an extra week because the gravity unexpectedly kept dropping after the second dry hop addition. They were at a loss to explain the behavior until hearing Shellhammer’s talk, after which everything made sense.

If any brewers out there have stories to share related to dry hopping and this research, feel free to share in the comments section here or on your preferred social media platform. If you are hungry for more details take a look at Tom Shellhammer’s research papers given in the reference section below.

Thanks to Mary MacDonald and Justin Hemminger of the Ohio Craft Brewers Association for allowing me to attend the conference as a member of the media.



[1] Not to be confused with the most enjoyable part of the conference, which was sampling barrel aged beers with Angelo Signorino and Dan Eaton at Barley’s Brewcadia on Wednesday evening. The Fate Barleywine was particularly spectacular.

[2] As long-time readers of this site will know, Cascade, the hop that launched the craft beer revolution in America, were developed at Oregon State University. See my earlier post on hops and the American IPA revolution for details.

[3] S. R. Lafontaine, T. H. Shellhammer, “Impact of static dry-hopping rate on the sensory and analytical profiles of beer” Journal of the Institute of Brewing 124, 434−442 (2018).

[4] S. Hieronymous, “For the Love of Hops” Brewers Publications, Boulder (2012).

[5] E. Parkin, T. H. Shellhammer, “Toward understand the bitterness of dry hopped beer” Journal of the American Society of Brewing Chemists 75, 363−368 (2017).

[6] K. R. Kirkpatrick, T. H. Shellhammer, “Evidence of Dextrin Hydrolyzing Enzymes in Cascade Hops (Humulus lupulus)” Journal of Agricultural and Food Chemistry 66, 9121−9126 (2018).

28 thoughts on “The Surprising Science of Dry Hopping – Lessons from Tom Shellhammer

Add yours

  1. When you speak about diminishing returns from Dry-hopping, did you interpret this information to be in total or for each dry hop round? For example, when double or triple dry-hopping, is it a waste to go over 1.1oz/gal for each dry hop charge or a waste to go over 1.1oz/gal for the whole batch?

    1. My interpretation is that this is the limit for all dry hop additions. If you saturate the solution on the first round of dry hopping, there wouldn’t be much gained by another round, solubity is solubility. Though that effect wasn’t specifically studied so maybe my instincts are wrong.

  2. I had massive problems with hop creep when I switched from my homebrew/test setup with Speidel plastic fermenters to conicals (to brew commerically).
    In the Speidels the same beer fermented to 3 °P, in the conical it never stopped fermenting after dry hopping and went down to 2,2-2,3 °P. Nice thin beer. I was totally desperate. I’m not sure what the difference is, maybe the pressure fermentation or the conical shape, but only dry hopped beers were affected.

    Then I read about the hop creep. Since then, I coldcrash in the conical after reaching the attenuation of the prior testbatch in the Speidel fermenter and keg. It does not continue to ferment after kegging, Co2 stays the same. I assume the enzymes are released over time because they work very quickly (as seen when mashing) and the yeast would not need 40 days to ferment 1.5°P. So when you separate the hop material form the beer, you seemingly stop the creep.

    Otherwise we would have thousands of exploding homebrewed IPA bottles, which are primed and bottle conditioned warm. Why some bottles or cans explode, I have no clue.

    1. I’m afraid not. Here is the exact protocol used for the dry hopping, taken directly from reference [3]. Maybe this will help you figure out how to do the extrapolation to T-90 pellets.

      To achieve the 200, 386, 800 and 1600 g hop/hL unhopped beer treatment rates, the whole cone hops were ground into a hop grist which was divided by mass into two mesh bags (EcoBag, Ossining, NY, USA). These bags were stored inside high barrier pouches flushed with nitrogen until dry-hopping. For each dry-hop treatment, the two kegs filled with 40 L beer were temporarily de-pressurised and opened under a stream of low pressure carbon dioxide. Simultaneously, the high-barrier pouch bag was opened and the mesh bag containing ground hop grist was added to the beer.

      Here they are grinding up the whole cone hops which I assume is similar to what is done when you make pellets. So my take is that the mass of whole cone and pellet hops would be similar. I could be wrong about that though.

  3. Would ABV be assume to take part in solubility of humulinones, so that higher ABV beers could take more dryhopping?

    1. Higher levels of ethanol would presumably lead to an increase in the solubility of the hydrophobic aromatic oils, but I don’t have a good feel for how large the effect would be. My first guess is that the differences would be subtle.

  4. So was it investigated that no Saccharomyces diastaticus was involved in the beers where the gravities dropped?

    1. There is no mention of saccharomyces diastaticus in the article describing the study. They did add sodium azide at the same time as the dry hops, which if I remember correctly was to kill any bacteria that might lead to additional fermentation.

    1. I don’t think the conclusions here are specific to lagers. Yes they used Coors Banquet beer for the experiment, but that beer doesn’t have any live yeast left in it (to my knowledge). The yeast that was added was California Ale yeast, Wyeast 1056.

      In the first part of the post the beer that was dry hopped was made according to the following specifications (taken directly from reference 3):
      The ‘unhopped’ wort was prepared with 86% pale two-row, 13.5% Caramel 10°L and 0.5% Caramel 120°L malt (Great Western, Vancouver, WA, USA) to a starting concentration of 11.3°P. Fermentation was performed using a Scottish ale yeast (Wyeast 1728) at 19.4–20°C. Following
      fermentation, a kieselguhr filter was used to clarify the green beer and remove yeast. Post filtration iso-humulones (IsoHop, John I Haas, Yakima,WA, USA) were added at a concentration of 18 mg/L. This resulted in ~55 hL of a 19.8 BU, 4.75% ABV ‘unhopped’ base beer.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

Blog at

Up ↑

%d bloggers like this: