wi n e maki n g

F I L T R A T I ON

Filtration of grape juice sediments:

a new application for cross-flow filters
By Patrik Schonenberger1, Davide De Giorgi1 and Julien Ducruet1,2
1
Engineering School of Enology in Changins, Route de Duillier 50, Case postale 1148, 1260 NYON 1, Switzerland
2
Corresponding author: Dr Julien Ducruet. Email: julien.ducruet@eichangins.ch
Introduction

rape juice for white wine is

usually pre-clarified to a certain
degree before undergoing
alcoholic fermentation. Traditionally,
this pre-clarification has been done by
allowing particles to settle overnight
and racking off the clear juice. The
settled particles (sediments) are
filtered with a filter press or a rotary
vacuum filter to obtain higher yields
of clear juice. These techniques have
proved to be efficient, but contain
inconveniences, such as the removal of
filtration residues (including adjuvants
like diatomaceous earth, perlite, etc.)
which may pose environmental concerns
(Montigaud 2010, Desenne 2004). Other
important concerns in wineries are the
initial dilution of the juice, little filter
versatility, and labour hours for the
cleaning procedure. Rotary vacuum
filters have been studied extensively
over the last 40 years. Some of the
main consequences found for the
filtered juices are linked to the use of
the vacuum itself: decrease of carbon
dioxide concentration and free sulfur
dioxide levels, losses of some desired
volatile compounds (e.g., primary
aromas), and increased risk for juice
oxidation (Riberau-Gayon et al. 1998).
The first cross-flow filters used
in the wine industry were reported
to have heavy impacts on the filtered
wines. Such wines reached lower
scores than wines made from rotary
vacuum-filtered juices, as evaluated by
professional tasting panels (Serrano
and Paezold 1998). In recent studies
with updated filter technology and
improved membrane characteristics,
no differences were found between
wines from cross-flow filtered and
rotary vacuum-filtered juices (Etienne
and Benestau 2000, Cunat et al. 2003).
However, Vernhet et al. (1998) analysed
wine components in the laboratory
and reported lower levels of colloidal
compounds in wines from cross-flow
filtered juices compared with wines from
rotary vacuum-filtered juices. Cunat et
al. (2003) noted in a study that ceramic

28

w w w.w i n eb i z .c om.au

Figure 1. Cross-flow filter FX 3 (Bucher-Vaslin SA, France), equipped with an

adapted module and the corresponding computer software, which allows the
filtration of very dense and viscous juice sediments.
membranes retained more colloids than
membranes made with polysulfone.
Cross-flow filters with adapted
modules for the filtration of dense
sediments appeared on the market
only recently. This new application for
cross-flow filters offers the possibility to
filter wines, juices, and juice sediments
with one device. The present study
was carried out as part of a Bachelor
of Sciences thesis (De Giorgi 2010) at
the School of Enology in Changins,
Switzerland, to evaluate the impact of
this new application on the quality and
composition of the final wines. Crossflow filtration of juice sediments was
compared with the standard filtration
with a rotary vacuum filter.
W i n e & V iticultur e Jo ur n a l MARC H /APRIL 2012

Materials and methods

Two experiments were carried out
with two pre-clarification methods
(gravitational settling and flotation).
After pre-clarification, sediments were
filtered with a rotary vacuum filter and
compared with the filtration from a
cross-flow filter.
Filter technology
A cross-flow filter FX 3 (Bucher
Vaslin SA, France) was equipped
with an adapted module and the
corresponding computer software,
which allows the filtration of very
dense and viscous juice sediments.
The membrane was made of polyether
V27N2

F I L T R A T I ON

sulfone with spaghetti-type capillaries

(internal capillary diameter 3mm).
The filtration surface was 18m2 with
0.2m pore size. Water (cold and hot)
and compressed air were connected
permanently for proper running of the
filter. The computer software controlled
all valve openings (wine, water and air)
and supervised filtration, rinsing, and
cleaning programming (Figure 1).
A rotary vacuum filter (VELO, Italy)
with an integrated vacuum pump and
a filtration surface of 8m2 was used
as the standard procedure. Filtration
media was Seitz Perlite A (Pall
SeitzSchenk Filter Systems GmbH,
Germany).
Grape juice
White Chasselas (Vitis vinifera L. cv
Chasselas) juice from a commercial
vineyard (Chteau dAuvernier,
Switzerland) was used. Enzymes
(10mg/L Depectil Clarification FCE,
Martin Vialatte) and SO2 (80mg/L) were
added to the juice before the application
of the treatments. The fraction of the
juice that was floated before filtration
received 200mg/L gelatin (proVgreen
Extra, Martin Vialatte) prior to
flotation, in addition to the enzymes and

wi n e maki n g

the SO2. No other product was added to

the juice. One fraction of the juice was
pre-clarified overnight by gravitational
settling at ambient temperature
(16C). A second fraction was preclarified by flotation, using a Turboflot
ECO 5000 (Kunzmann and Hartmann,
Germany) with an average flow rate of
7500L/h. Flotation gas was nitrogen.
The sediments of both pre-clarification
methods were then filtered using both
cross-flow and rotary vacuum filtration.
Inherent to rotary vacuum filtration, the
first fraction of the juice was diluted
and not used for the winemaking.

separately with 0.2g/L Vitilevure Quartz

(Saccharomyces cerevisiae galactose,
Station Oenotechnique de Champagne,
France). For the malolactic
fermentation, 0.01g/L Vitilactic F
(Martin Vialatte, France) was added to
the wines. SO2 at a rate of 50mg/L was
added after malolactic fermentation,
and the wines were then stored at 2C
for a period of six weeks for physical
stabilisation. Wines were pre-filtered
with AF100 filter pads (Filtrox AG,
Switzerland) and, before bottling, were
sterile filtered with AF130 filter pads
(Filtrox AG, Switzerland).

Winemaking
Special care was taken to use only
juice fractions with the same sugar
levels (total soluble solids) before
and after filtration. After filtration,
wines were replicated three times in
stainless steel tanks (each containing
200L of juice). Standard procedures
for white Chasselas wine production in
Switzerland were applied, and attempts
were made to keep the temperatures
identical in all tanks, i.e., alcoholic
fermentation at 18C and malolactic
fermentation at 16C. For the alcoholic
fermentation, all tanks were inoculated

Chemical analysis
The sugar content of the juice
was measured with a standard
refractometer. Alcohol content,
titratable acidity, tartaric acid, malic
acid, lactic acid, and volatile acidity
were analysed with a WineScanTM (FOSS
Analytical, Hilleroed, Denmark) which
uses Fourier Transform Infrared (FTIR)
Spectroscopy. The FTIR analysis was
carried out with the clear fraction of
the juice after centrifuging the samples
for 10 minutes at 5000rpm. Samples
were also centrifuged for calculating
the proportion of dry matter in the total

Winemakers bottling
for winemakers
With ten winemakers working across six sites, Portavin
is close to market and transport hubs, saving time,
money and the environment.
Portavin caring for your wine from tank to shelf

Adelaide
Auckland
(08) 8447 7555 (09) 582 0090

Margaret River
(08) 9755 0500

Melbourne
(03) 9584 7344

Perth
Sydney
(08) 9437 1033 (02) 9722 9400

www.portavin.com.au
portavin@portavin.com.au

V2 7N 2

W i n e & V iticultur e Jo ur n a l MARC H /APRIL 2012

www.w inebiz. com . au

29

wi n e maki n g

F I L T R A T I ON

Table 1. Chemical analyses of the initial juice after pre-clarification by gravitational settling and flotation.
Juice analysis

total soluble solids (Brix)

Titratable acidity (g/L)

pH

Dry matter (% of total volume)

After gravitational settling

19.1

7.10

3.29

8.78

After flotation

19.1

7.30

3.22

28.25

volume. The volume of the centrifuge

deposit was divided by the initial total
volume to obtain the percentage of dry
matter.
Statistical analysis
Data were subjected to the analysis
of variance (ANOVA) and F-test. The coefficient of variation (CV) was calculated
to obtain the relative percentage of
the standard deviation (Excel 2007,
Microsoft, Redmond, WA).
Sensory evaluation
The bottled wines were stored for
three months before being exposed
to sensory evaluation by an expert
panel. Two panels of 24 and 16
judges, respectively, were present
at two tasting sessions. Quantitative
Descriptor Analysis profile (QDA) with
10 different terms was employed to
describe the wines.

Results and discussion

Chemical analyses of the initial
juices are shown in Table 1 (total
soluble solids, titratable acidity, pH,
and dry matter as a percentage of
the total volume). The two types of
sediments obtained by flotation and
gravitational settling were distinctly
different: sediments from flotation
were denser and more viscous than
sediments from gravitational settling
(28.25% versus 8.78% dry matter of
total volume, Table 1). Both filtration
techniques cleared the juice to a
very low but comparable level of
suspended material (<1.00% dry matter
of total volume, data not shown). A
temperature rise of 6C was observed
in the juice during cross-flow filtration
(data not shown) and needs some
consideration. Similar differences for
input and output temperatures for
cross-flow filtered wines have been

reported previously in comparative

studies (Ducruet et al. 2006). Cross-flow
filtration may negatively affect a juice
with a higher input temperature. It is
preferable to conduct the filtration with
an initial juice temperature below 16C
to reach a maximum temperature of
22C at the output side of the filter.
The chemical analyses of the finished
wines showed no significant differences
for alcohol content, acidity levels, and
volatile acidity (Tables 2 and 3).
Significantly lower absorbance at
280nm was observed in wines made
from rotary vacuum-filtered juices
compared with those from cross-flow
filtered juices (Tables 2 and 3). The
absorbance at 280nm is an indication
for total polyphenols in the juice. If
rotary vacuum filtration caused more
polyphenol oxidation than crossflow filtration (Riberau-Gayon et al.
1998), one portion of these oxidised

Filtration & Purification Specialists

For all of your winery and beverage processing needs.
Premium quality filtration solutions and specialists you can trust.
Complete filtration solutions from water to wine
Premium BECO sheets & lenticular modules
Turnkey filtration systems
Australias largest range of filter housings
SIHA biotechnology & wine additives

H 2O

30

w w w.w i n eb i z .c om.au

CALL FOR MORE INFORMATION

Blue H2O Filtration Pty Ltd
29 Dalgety Street Oakleigh VIC 3166
Ph +61 3 9564 7029 Fax +61 3 9564 7039
info@blueh2o.com.au www.blueh2o.com.au

W i n e & V iticultur e Jo ur n a l MARC H /APRIL 2012

V27N2

F I L T R A T I ON

wi n e maki n g

Table 2. Chemical analyses of the final wines made from juice obtained by gravitational settling. Juice was filtered, vinified,
stabilised, and stored for three months in the bottle.
Wine
analysis

Alcohol
(% volume)

Tartatic acid
(g/L)

Malic acid
(g/L)

Lactic acid
(g/L)

Acetic acid
(g/L)

Volatile
acidity (g/L)

Absorbance (280 nm)

Chromtic intensity
(absorbance 420 nm)

Juice obtained by gravitational settling

Rotary
vacuum
filter

10.9

1.73

0.41

1.94

0.23

0.42

5.267

0.143

Cross-flow
filter

11.1

1.70

0.47

2.25

0.18

0.42

6.000

0.154

significance
CV %

ns

ns

ns

ns

ns

ns

**

ns

3.05

3.62

4.45

3.77

21.25

3.09

2.33

19.24

*, **, ***, ns: Main effects significant at P < 0.05, P < 0.01, P < 0.001, or not significant, respectively.
Table 3. Chemical analyses of the final wines made from juice obtained by flotation. Juice was filtered, vinified, stabilised, and stored
for three months in the bottle.
Wine
analysis

Alcohol
(% volume)

Tartaric acid
(g/L)

Malic acid
(g/L)

Lactic acid
(g/L)

Acetic acid
(g/L)

Volatile
acidity (g/L)

Absorbance (280 nm)

Chromatic intensity
(absorbance 420 nm)

Juice obtained by flotation

Rotary
vacuum
filter

11.3

1.67

0.48

2.05

0.28

0.47

5.650

0.223

Cross-flow
filter

11.2

1.65

0.51

2.18

0.13

0.43

6.310

0.126

Significance

ns

ns

ns

ns

**

ns

0.04

0.04

0.22

4.73

18.62

8.51

7.52

45.77

CV %

*, **, ***, ns: Main effects significant at P < 0.05, P < 0.01, P < 0.001, or not significant, respectively.

FILCHEM
A Complete Range
of Filter Media
Ertel Alsop lenticular
cartridges & sheets
Graver membrane
cartridges
Celite DE
Harborlite Perlite
Diacel cellulose
Activated Carbon
Filchem Australia Pty. Ltd

1 800 33 1125
V2 7N 2

Fast and ACCURATE results for wine analysis with the

Thermo range of Gallery and Arena Discrete Analysers
Compact design occupies a small footprint
and is fully self-contained.
Flexible loading capacity up to 45 samples
or 30 reagents simultaneously.
All necessary steps are automated, providing
a walk-away time up to two hours.
For more information on this product,
contact us by email at sales.
daniel.hoger@thermofisher.com
www.thermofisher.com.au
1800 333 110

Thermo Scientific Gallery

Photometric Analyzer

Moving science forward

W i n e & V iticultur e Jo ur n a l MARC H /APRIL 2012

www.w inebiz. com . au

31

wi n e maki n g

F I L T R A T I ON

suggested that other factors than the

application of the treatments may have
influenced these data (Table 3).
The QDA profiles of the tasting panels
showed few differences in the final
wines (Figures 2 and 3). Cross-flow
filtration of juice sediments produced
wines with less intense odours and more
pronounced acidity levels compared with
wines from rotary vacuum-filtered juice
sediments.
Conclusion

Figure 2. Quantitative Descriptor Analysis profile (QDA) of the final wines made
with juice obtained by gravitational settling (*, **, ***: Main effects significant at
P < 0.05, P < 0.01, P < 0.001).
polyphenols might have flocculated
and immediately filtered out of the
juice, which did lower total polyphenols
in rotary vacuum-filtered juices. The
portion of the remaining oxidised
polyphenols (indicated by chromatic
intensity) was not different in both
filtration techniques, if the juice was

obtained by gravitational settling (Table

2). On the other hand, if the juice was
obtained by flotation, the portion of
the oxidised polyphenols (indicated by
chromatic intensity) and acetic acids
were increased in rotary vacuumfiltered juices (Table 3). However, a high
CV of 18.62% and 33.3%, respectively,

MaloBacti MLF

Cross-flow filters offer several

possibilities to facilitate cellar work.
They produce less filtration residues
and may free labour hours through
automation during intense harvest time.
Cross-flow filters are more versatile
than rotary vacuum filters; one filter
may be used for wine, juice, and juice
sediment filtration. However, initial
investments are more substantial for
a cross-flow filter. Some products
(e.g., bentonite, activated charcoal)
may not be added to the juice before
cross-flow filtration, in order to prevent
the membrane from plugging. These
experiments were carried out with
one grape variety and one vintage.

Innovation.
Performance.
Proximity.
Better equipment
is the key to success
Delta

Fruit receival

MAXIMUM SECURITY MALOs

The only strains on the market with a pH test that
confirms viability of active bacteria prior to inoculation!

Bucher
Presses

Flavy

MaloBacti CN1

Cross flow filtration

Citric negative MLF

Flavy

MaloBacti HF2

Cross ow ltration

Colour protection & flavour management

Our dealership network Australia and New Zealand

MaloBacti AF3
Elevated phenolics & high alcohol

MAXBacti
New large volume concept

MaloControl

SALES NZ
Viniquip

SALES VIC/TAS/ACT
Vinvicta

Horst Klos
Steve Jenkinson
Our dealership network Australia and
New Zealand
Tel: (06) 879 7799
T: 1300 360 353

SALES & SERVICE WA

SWAT trading
Rob Menzies
T: (08) 9755 5766

South Australia and

Western Australia:
SWAT Trading
Tel: 08 9755 5766

New
South WalesSERVICE
and & PARTS
SALES SA
Filters SA
APV Australia
Queensland:
T: (08) 8388 3999
Les Jarvis
Wine Energy
T: 1800 100 278
Tel: 0407 400 728

Victoria and Tasmania:

Vinvicta Products
Tel: 03 9464 7414

Po Box
1051, Glen Waverley, 3150, tel 1300 Bucher
New
Zealand:
National Sales
Manager: Peter Keeghan
Viniquip
International:
tel.
0417
816 024
Email: peter.keeghan@buchervaslin.com
Tel: +64 9578
3740

SERVICE & PARTS

APV New Zealand
Grant Dewson
T: 800 500 278

Bucher Vaslin Australia

Complete MLF supplement

For further information, please contact Kauri NZ Ltd.
NZ Tel: 0800 KAURIWINE AUS Tel: 1800 127 611
NZ Fax: 04 910 7415
AUS Fax: 1800 127 609
Email: winery@kauri.co.nz Web: www.kauriwine.com

32

w w w.w i n eb i z .c om.au

Bucher Vaslin Australia Pty Ltd

PO Box 1051,
Glen Waverley, 3150
Samuel Plumejeau
Tel: 0427 655 800

W i n e & V iticultur e Jo ur n a l MARC H /APRIL 2012

www.buchervaslin.com
Seesuccess
us atis our
Winetech
Your
priority
July 29th to August 1st in Adelaide

www.buchervaslin.com
Your success is our priority
V27N2

F I L T R A T I ON

wi n e maki n g

Further studies need to be done to better

understand the behaviour of the juices
and finished wines over time after using
cross-flow filter technology.
Acknowledgements
The project was conducted as part of
Davide De Georgis Bachelor of Science
thesis, which was supervised by Dr
Julien Ducruet. We greatly appreciated
the technical assistance of M. Droz
Frdrique (Chteau dAuvernier,
Switzerland) and M. Erik Dobrovolski
(Bucher Vaslin SA, France).
References
Cunat, P.; Lorenzini, F. and Bregy, C.A. (2003)
Comparaison de membranes en cramique et
polysulfone pour la microfiltration tangentielle des
vins. Revue Suisse de Viticulture, Arboriculture,
Horticulture 35(6):110-119.
DeGiorgi, D. (2010) Etude de la filtration
tangentielle des bourbes compare aux techniques
existantes. BS Thesis, Engineering School of Enology
in Changins, Switzerland.
Desenne, A. (2004) Les filtrations: une
pollution diffrente selon le typ de filtre. Chambre
dagriculture de la Gironde service vigne et vin.
http://www.matevi-france.com/welcome_menu.
asp?tp=choix_experimentation [accessed 10/02/2010]
Ducruet, J.; Silvestri, A.-C. and Hyppenmeyer,
P. (2006) Etude comparative de diffrents filtres

tangentiels en nologie. Revue Suisse de Viticulture,

Arboriculture, Horticulture 38(5):297-302.

Chimie du vin stabilisation et traitements. Edition

Dunod: 383-427.

Etienne, F. and Benestau, F. (2000) Filtration

tangentielle: impact sur la qualit des vins. Revue
des nologues 27(96):13-15.

Serrano, M. and Paetzold, M. (1998) Incidence des

filtrations sur la composition chimique et les qualits
organoleptiques des vins. Journal International des
sciences de la vigne et du vin. Traitements physiques des
mots et des vins. Filtration. N hors srie. pp 53-57.

Montigaud, I. (2010) Les terres de filtration

pourraient finir en compost. Russir Vigne. N164:
35.
Riberau-Gayon, P.; Glories, Y.; Maujan, A. and
Dubourdieu, D. (1998) La clarification des vins par
filtration et centrifugation. Trait dnologie. Vol 2.

Conventional

WIN
EB
AR
of
ort
nsp
tra
nd
ga
lin

ACKS
LR
RE

Flexibil
ity f
or s
tac
kin
g,
ha
nd

Figure 3. Quantitative Descriptor Analysis profile (QDA) of the final wines made
with juice obtained by flotation (*, **, ***: Main effects significant at P < 0.05,
P < 0.01, P < 0.001).

arrels
wine b

Proudly designed and manufactured by

JOHN FALLAND AUSTRALIA
Setting Standards Since 1961
Moppa Road South, Nuriootpa, SA 5355
Ph: (08) 8562 1533 Fax: (08) 8562 2103
Email: john@jfallandaust.com.au
Website: www.jfallandaust.com.au

Fabricated 20 years
ago and still popular
Over 270,000
made for over 1200
wineries
Stack five high with
your forklift
Lifetime galvanized
finish
Competitive prices
Reg Design No
117931

Vernet, A.; Moutounet, M. and Escudier, J-L.

(1998) Microfiltration tangentielle des vins. Journal
international des sciences de la vigne et du vin.
Traitements physiques des mots et des vins. Filtration.
N hors srie 45-52.
WVJ

or

Barrel Master

Superior rack
on rack stack
system
Simple,
uncluttered
design
Visible & easy
rack locating
system
No weight on
barrels
Static or barrel
rolling options
Reg Design
No 154262

New Zealand sales and distribution:

KAURI NEW ZEALAND LIMITED
Ph: 04 476 0105 Fax: 04 476 0161
PO Box 17-385,
. Karori, Wellington.

V2 7N 2

W i n e & V iticultur e Jo ur n a l MARC H /APRIL 2012

www.w inebiz. com . au

33