Automation in the Wine Laboratory
The introduction of automated methods of testing in wine laboratories has been ongoing for many years. In this article the tests that can be easily automated and the types of instruments available are discussed. The benefits and the problems of installing automated equipment are also outlined.
One of the main aims of automating testing in the wine laboratory is to be able to perform a larger volume of tests in a shorter period of time. The desire to reduce operating costs, particularly labour, is also important. One other objective is to make the results of the tests more repeatable. These underlying reasons will be considered below.
Benefits of automating testing
By automating lab testing a larger number of tests can be done on any day by fewer staff. Automated instruments do not need to be used all day every day.
The reliability of the test results can also be increased by using automated instruments as the amount of manual manipulation is reduced, thus random errors due to human causes can be decreased.
The quantity of sample and reagents needed to do the tests is usually much lower and this can also be a cost saving, particularly where the costs of reagents are high. One area that has high reagent costs is enzymatic testing done by a spectrophotometer. This is a very common technique in wine labs and the enzyme reagents are expensive; by reducing the amount of reagents needed considerable costs can be saved.
Tests that can be automated
Some tests are more easily automated than others. A look through some larger wine laboratories will usually show one or more of the tests listed below having some level of automation. It is important to realize that there is no one answer to this situation and it will depend upon the type of wine made, the experience of the laboratory staff, and the desire of the management of the enterprise to invest in the capability of the laboratory. The instruments listed below are in the order in which it is considered they would be added.
These tests are ones that are commonly automated in Australian and New Zealand wine labs:
pH and Titratable acidity (TA)
These tests can be very time consuming when done manually. For TA in particular, the manual steps of degassing samples and then titrating are quite laborious. Automating both degassing and titrating is very easy when using an autotitrator. These are simple devices to use and can be the cheapest step in the road to automating the laboratory.
Autotitrators can have various configurations. The degassing function is a very cheap option on some models and a great labour-saving device: autotitrators that do not degas the sample are only doing half the job. In our labs they are considered to be an essential option with any autotitrator.
Figure 1 Autotitrator without sample changer
A more expensive option is a sample changer. This allows a number of samples to be analysed in the same run, instead of just one sample being presented to the instrument at a time. Sample changers are expensive however and can more than double the original investment. They can be added as an option at a later date if it is found that the volume of samples increases and the extra automation is deemed to be warranted.
Figure 2 Autotitrator with sample changer
Alcohol in wine can be measured in various ways. The boiling point depression technique using an ebulliometer is simple and easy to use but is not accurate enough for wines that are destined for export, although it can be used with some caution for domestic labeling purposes.
The manual methods of measuring alcohol accurately involve a distillation step. This is time consuming, uses a large amount of sample and requires a lot of operator attention. The distillate is then used to determine that alcohol content of the wine by either hydrometry or a weighing vessel (pycnometer). Both techniques are delicate and require very good operator skills, strict temperature control, and in pycnometry an accurate 4 decimal place balance. The distillation/pycnometry technique is very accurate and is used as the international reference method.
The distillation step in these methods can be automated to varying degrees, but this still leaves a manual step for the final measurement. We have not gone down this automation path in our labs because of the advent of the next instrument:
Near infra red (NIR) spectrophotometers are the simplest and most accurate instrument for measuring alcohol; not surprisingly they are also the dearest instrument for performing alcohol tests.
Sole purpose alcohol NIR spectrophotometers are very popular for alcohol testing in wine as they require little calibration and only take only a minute or two per sample.
Figure 3 Alcohol NIR spectrometer
Larger NIR instruments that can perform multiple analyses simultaneously are also available. These can measure alcohol, pH, TA, malic acid etc. in the one scan.
These instruments are not necessarily very accurate for all analytes and can also suffer from serious matrix effects. They also need considerable “calibration” that may take months or years to achieve. A few of these are in use in Australia and NZ, and if their limitations are properly realized and managed, they can be useful instruments in certain circumstances.
Enzymatic tests for wine encompass several analytes such as glucose and fructose (residual sugars), malic acid, acetic acid (volatile acidity), YAN (yeast assimilible nitrogen).
Simple spectrophotometers are the most popular instrument in small to medium sized wine labs after pH meters. They are quite simple to use, are very robust, and usually require little maintenance. Their main drawbacks of using these instruments are: expensive reagent use due to large cuvette size (3mL) and manual effort. Spectrophotometers can be purchased that hold 1, 4 or 6 cuvettes at a time; the more that can be done in one run, the less effort needed for taking readings.
There are several options for automating spectrophotometers and as expected, the better the system the more expensive the price.
These are the cheapest automated photometric instrument but also require the most manual effort to operate them. Instead of single cuvettes they use plates that have 96 wells so they can do many samples in the one run. Each well holds about 0.3mL of solution, thus using less reagents compared with a 3mL cuvette.
Microplate readers still require all samples and reagents to be introduced by micropipettes. Multi channel micropipettes that have 8 tips can be used to reduce the manual work. Once the reagents are pipetted into the 96 plate well, the rest of the process can be automated including data retrieval by a PC and the calculations performed by an accompanying program.
Figure 4 Microplate reader for enzymatic analysis
There are several brands and configuration of autoanalysers that can be used for wine analysis. The common principles of operation include:
- all reagents are placed in special containers in the analyzer itself
- an automated pipette draws up and dispenses reagents in the required amounts and sequence
- in some instruments multiple tests can be done on the same sample in the same run
- measurement by ultra violet or visible light source (UV/Vis)
- microcuvettes used and so very low amounts of reagents used
- PC driven with all readings and calculations performed automatically by a preset program.
Figure 5 Autoanalyser for enzymatic analysis
One area of concern with some autoanalysers is that they are set up for using only proprietary brand reagents. As all tests use enzymatic reagents this can be limiting and costly. With some brands the supply of reagent kits can be sporadic thus causing major disruptions during peak periods. Autoanalysers that can be programmed to work with any brand of enzyme reagent are much more flexible and less troublesome.
The instruments mentioned above are offered by a number of manufacturers and their distributors. Some of these suppliers can be found in the Wine Industry Directory Buyers Guide in the Laboratory Equipment section.
Costs of automation
The main cost of automating a wine lab is the capital cost of the equipment. As well as this most of these instruments require a specialist technician to come on site for the initial installation plus set up and training of staff. This entails extra costs on top of the initial capital outlay.
Continual calibration of the instruments is an extra regular task that creates more labour compared to manual methods and in some instances requires the purchase of calibration equipment.
Below is a table giving some price ranges for the different type of instruments. As there are many brands and models on the market, this is designed to give a rough guide only.
|Automated instrument||Approximate price range (A$000)|
|Autotitrator||$5 - $25|
|NIR spectrophotometer||$25 - $150|
|UV/Vis spectrophotometer||$2 - $8|
|Microplate reader||$20 - $25|
|Autoanalyser||$50 - $100|
The major drawback to installing automated instruments is the much larger capital cost compared to traditional wet chemistry techniques. The level of expertise of lab personnel also needs to be higher and this may mean that higher wages need to be paid as well
Maintenance costs are high when required, however many automated instruments can run trouble free for extended periods of time. If breakdowns happen in very busy periods, such as vintage, this can cause quite stressful situations as the lab workflow can be seriously disrupted until repairs are carried out. As many winery labs are located in rural areas the availability of specialist technicians for troubleshooting can be both costly and suffer from delays.
Although reagent costs can be much lower, some other consumables, such as special cuvettes or vials, can be quite expensive. Generally though, savings are made on consumable costs.
Using automated equipment can certainly reduce labour and consumable costs and produce a much higher throughput of samples than in a manually operated laboratory. The savings made need to be compared to the cost of capital for the initial purchase of the equipment. A cost benefit analysis should be made for each purchase.
Copyright© 2011 Vintessential Laboratories. No part of this publication may be reproduced in any form or by any means without the permission in writing from the copyright owner.