BarthHaas UK
BarthHaas UK specialises in producing a wide range of natural ingredients derived from super-critical CO2 hop extracts. These hop extracts contain all the alpha-acids, beta-acids, waxes, resins and the essential oils found in hops, without the vegetative material and almost no loss of flavour characteristics. This means that CO2 extracts can be used to partially or entirely replace leaf hops or hop pellets in the brewing process, providing added efficiency and flexibility for brewers.
Oil-enriched CO2 Hop Extract
Oil-enriched CO2 Hop Extract is a variant of CO2 hop extract that has a higher concentration of hop oils, approximately 15 – 40 mL/100 g, depending on the variety. The color of the extract can vary from yellow to dark green. It is a semi-fluid paste at room temperature, becoming more fluid when heated.
Traditional Laboratory Methods
Traditional methods for measuring alpha and beta acids in hop extracts involve the use of toxic chemicals such as lead acetate (EBC 7.7 or ASBC-Hops14), methanol and diethyl ether (EBC 7.6 or ASBC Hops-8). These reagents are expensive, corrosive and require hazardous waste disposal processes.
The traditional method for measuring oil content by steam distillation (EBC 7.10 or ASBC Hops-13) does not involve hazardous chemicals. However it is a very slow and laborious method, taking up to 4 hours per sample.
FTNIR Spectroscopy
As an alternative to all these methods, FT-NIR spectroscopy can provide rapid quantification without the need for solvents and reagents. Using reference values collected using the traditional analysis methods, FT-NIR calibration models can be trained to simultaneously determine the alpha, beta and oil content of unknown CO2 hop extract samples
- no hazardous chemicals
- in seconds, not hours.
Calibration Models With Machine Learning
Sagitto's fully-managed calibration model building service ensures that users of FT-NIR instruments such as BarthHaas UK can quickly get the most from their investment.
Once an initial calibration model is built using a small collection of FT-NIR spectra and reference values, a process of continuous improvement starts. Using appropriate cross-validation techniques to assess the model’s accuracy, outliers can be identified and removed. As more data is collected, to fill gaps and extend the range of the values being predicted, the model is rebuilt and new cross-validation statistics are generated.
Benefits to BarthHaas UK
Sagitto has built more than 45 calibration models for BarthHaas UK, allowing them to analyse a wide range of their products using FT-NIR with its corresponding speed and productivity benefits.
"Using FT-NIR and models built by Sagitto, we can save around three hours per batch. This means that in a year we could be saving up to a month in production time just for one product, allowing us to make many more batches."