Hildegarde Heymann, and David E. Block

5/17/2019

1

Peter Buffon, Luke Bohanan, Jeff Strekas, Hildegarde Heymann, and David E. BlockDepartment of Viticulture and EnologyUniversity of California, Davis

“Some wine makers, especially those at smaller, artisanal wineries, believe that this process may strip subtle aromas and flavors from the finished wine, along with the haze, and so they make their wines by more traditional processes. The clarity may suffer (although most well-made wines come out clear even without filtering), but these wines are certainly as good, and perhaps even better, than the industrial-type wines.”--wine lovers page

The easiest way to tell whether a wine has been filtered is to observe its level of clarity. An unfiltered red wine may appear murky, "polluted" or dark. This is not a bad thing, it simply means that the wine was probably not filtered. Like coffee, wine is put through a filtration process that removes small particles. Many winemakers argue, however, that this process is not only unnecessary but actually bad for the wine as filtration can affect the wine's flavor and aroma. Read more: What Is an Unfiltered Red Wine? | eHow.com

However, these filtering processes may also remove elements that affect the flavors and aromas of a wine, so some winemakers choose not to filter. They believe that filtering strips the wine of its true character, and employ other methods of getting the wine as clear as possible (racking, cold stabilization, and other old-school techniques are the alternative). There are also wineries that avoid filtering — or keep it to an absolute minimum — to maintain organic status.--wine weekly

Some wine makers believe that this process may strip subtle aromas and flavors from the finished wine

Many winemakers argue, however, that this process is not only unnecessary but actually bad for the wine as filtration can affect the wine's flavor and aroma.

some winemakers choose not to filter. They believe that filtering strips the wine of its true character

Clarity

Microbial Stability

Need a systematic study to examine the effects of filtration

5/17/2019

2

Discuss common wine filtration processes Investigate any transient changes in chemical composition during filtration

Evaluate the sensory and chemical impact of pad sterile cartridge filtration on red and white wines

Evaluate the effects of cross‐flow filtration Evaluate the effects of pumping

Millipore

Housing

Filter

Pad or Plate and Frame Filter

Cartridge or Sterile Filter

5/17/2019

3

DepthProbability of hitting fiber

MembraneUniform pore size/size exclusion

Cabernet Sauvignon

Control – No Filtration

Filtration Through Empty Housing

Filtration Through 0.45 µm PVDF Filter

Filtration Through 0.45 µm PES Filter

Chemical AnalysisSensory Analysis

Time Points:0 Weeks3 Week5 Weeks7 Weeks9 Weeks…

Rep

 1

Rep

 1

Rep

 2

Rep

 3

Rep

 1

Rep

 2

Rep

 3

Rep

 1

Rep

 2

Rep

 3

• Push with nitrogen (no pump)• Filter right into bottling line• Extended one run of PVDF to

look at transient behavior

• 2007 Sonoma Valley• Post ML, oak aging

5/17/2019

4

Bottle Number

0 20 40 60 80 100 120 140

Col

or P

aram

eter

0.0

0.2

0.4

0.6

0.8

1.0

A420 A520 Hue Density

Bottle Number

0 20 40 60 80 100 120 140

Ta

nnin

(m

g ca

taec

hin

eq/L

)

0

50

100

150

200

250

300

Figure 2. Tannin as a function of bottle number. Bottles were sampledfrom the bottling line during the course of filtration with a PVDF membranefilter. Tannin was measured on a sample from each bottle using theAdams-Harbertson Assay.

5/17/2019

5

Not much difference during filtration—Now let’s examine differences between the filtration treatments

, , pStandard Composition

Bitter 0.75 g caffeine*Sour 1.5 g citric acid*Sweet 3.5 g sucrose*Astringent 312 mg alum*High-Viscocity 30 g/L Polycose*Low-Viscocity H2OBerry/Currant 1 sliced strawberry, 3 sliced raspberry, 3 sliced blackberry, 5 mL cassis**Cherry 10 canned bing cherries, 3 tsp cherry juice, 2 tsp cherry pie filling**Dried Fruit 15 raisins, 2 sliced prunes, 1 sliced apricot**

Vegetal1 cm2 green bell pepper, 2 cut string beans, 1 tsp canned asparagus juice, 1 tsp canned green bean juice, 5 blades grass**

Spice 1/8 tsp cinnamon, 1/8 tsp all spice, 3 cloves**Black Pepper 1/8 tsp ground black pepper**

Floral2.5 mL of 1 drop rose essence in 200 mL H2O, 2.5 mL of 1 drop violet essence in 200 mL H2O**

Leather/Smoke 3x1 cm pieces of leather shoe lace, 1/8 tsp liquid smoke**Solvent/Chemical 1 drop nail polish remover in 10 mL H2O**Oak 3 small American oak chips**Vanilla 5 mL Vanilla flavoring**Chocolate 1.5 chopped chocolate chips**Hot/Ethanol 15 mL vodka**

* added to 1L H2O** added to 50 mL Franzia Cabernet Sauvignon

5/17/2019

6

Pr>F Value Time Time*Filter Time*Rep Time*JudgeBerry/Currant 0.0835 0.1272 0.0163* <0.0001*Cherry <0.0001* 0.2252 0.9941 <0.0001*Dried Fruit <0.0001* 0.5785 0.8705 <0.0001*Vegetal <0.0001* 0.6287 0.2929 <0.0001*Spice 0.0095* 0.7531 0.936 <0.0001*Black Pepper 0.0016* 0.9479 0.3753 <0.0001*Floral <0.0001* 0.5798 0.4427 <0.0001*Leather/Smoke <0.0001* 0.8235 0.0278* <0.0001*Solvent/Chemical <0.0001* 0.4957 0.5963 <0.0001*Oak <0.0001* 0.0191* 0.3618 <0.0001*Vanilla 0.0003* 0.3555 0.9706 <0.0001*Chocolate 0.0016* 0.8618 0.5329 <0.0001*Hot/Ethanol <0.0001* 0.9904 0.8951 <0.0001*Astringent <0.0001* 0.1419 0.5326 <0.0001*Bitter <0.0001* 0.4845 0.7347 <0.0001*Sour <0.0001* 0.8656 0.923 <0.0001*Viscous 0.0014* 0.4533 0.1336 <0.0001** Value is significant at alpha = 0.05

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

Berry

Cherry

DrFruit

Vegetal

Spice

Pepper

Floral

Leath/Sm

SolventOak

Vanilla

Choc

EtOH

Bitter

Astrin

Sour

Visc

CTL W0

EMP W0

PES W0

PVD W0

0 weeks

5/17/2019

7

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

Berry

Cherry

DrFruit

Vegetal

Spice

Pepper

Floral

Leath/Sm

SolventOak

Vanilla

Choc

EtOH

Bitter

Astrin

Sour

Visc

CTL W0

EMP W0

PES W0

PVD W0

0

0.5

1

1.5

2

2.5

3

3.5

4

Berry

Cherry

DrFruit

Vegetal

Spice

Pepper

Floral

Leath/Sm

SolventOak

Vanilla

Choc

EtOH

Bitter

Astrin

Sour

Visc

CTL W3

EMP W3

PES W3

PVD W3

0

0.5

1

1.5

2

2.5

3

3.5

4

Berry

Cherry

DrFruit

Vegetal

Spice

Pepper

Floral

Leath/Sm

SolventOak

Vanilla

Choc

EtOH

Bitter

Astrin

Sour

Visc

CTL W5

EMP W5

PES W5

PVD W5

0

0.5

1

1.5

2

2.5

3

3.5

4

Berry

Cherry

DrFruit

Vegetal

Spice

Pepper

Floral

Leath/Sm

SolventOak

Vanilla

Choc

EtOH

Bitter

Astrin

Sour

Visc

CTL W7

EMP W7

PES W7

PVD W7

0

0.5

1

1.5

2

2.5

3

3.5

4

Berry

Cherry

DrFruit

Vegetal

Spice

Pepper

Floral

Leath/Sm

SolventOak

Vanilla

Choc

EtOH

Bitter

Astrin

Sour

Visc

CTL W9

EMP W9

PES W9

PVD W9

0 weeks 3 weeks 5 weeks

9 weeks7 weeks

A420* A520* Color Hue* Color Density*

Least Significant Difference 0.0491 0.0567 0.0043 0.1058A

Day 1 PVDF 0.3841A 0.4455A 0.8616A 0.8297A

Day 3 PVDF 0.3510A 0.4077A 0.8610A 0.7587A

Day 2 PES 0.3490A 0.4053A 0.8610A 0.7543A

Day 1 PES 0.3670A 0.4267A 0.8602AB 0.7937A

Day 2 PVDF 0.3790A 0.4408A 0.8599AB 0.8198A

Day 3 PES 0.3765A 0.4382A 0.8596AB 0.8147A

Day 3 No Filter 0.3516A 0.4092A 0.8596AB 0.7608A

Day 1 No Filter 0.3791A 0.4412A 0.8594AB 0.8203A

Day 1 Control 0.3510A 0.4097A 0.8566B 0.7607A

Day 2 No Filter 0.3633A 0.4243A 0.8562B 0.7877A

* Different superscripts denote a significant difference at alpha = 0.05Each reported value is an average of six replicates

5/17/2019

8

Tannin*mg catechin eq/L

Least Significant Difference 20.868

Day 1 PES 232.42A

Day 2 PES 231.03A

Day 1 No Filter 230.84A

Day 2 PVDF 230.58A

Day 3 No Filter 230.35A

Day 1 Control 230.16A

Day 1 PVDF 227.20A

Day 3 PES 226.56A

Day 3 PVDF 225.58A

Day 2 No Filter 217.85A

*Different superscripts denote a significant difference at alpha = 0.05Each reported value is an average of six reps

Merlot

Control – No Filtration

Filtration Through Pad 

Filter

Filtration Through Pad and 1 µm Depth Filter

Filtration Through Pad, 1 µm Depth, and 0.45 µm PVDF Filter

Time Points:1Weeks2Week4Weeks6Weeks12Weeks16 Weeks…

Rep

 1

Rep

 1

Rep

 2

Rep

 3

Rep

 1

Rep

 2

Rep

 3

Rep

 1

Rep

 2

Rep

 3

Filtration Through Pad, 1 µm Depth, and 0.45 µm PES 

Filter

Rep

 1

Rep

 2

Rep

 3

Chemical AnalysisSensory Analysis

• 2009 Oakville• Post ML, oak aging

• Push with nitrogen (no pump)• Filter right into bottling line

5/17/2019

9

Judge Filter Bottling Day

Time Filter* Bottling

Filter* Time

Filter* Judge

Mixed Berry <.0001* 0.2297 0.5788 <.0001* 0.9711 0.8974 0.9309

Medicinal Cherry  <.0001* 0.2331 0.2599 0.0002* 0.0993 0.4004 0.9711

Fresh Vegetable <.0001* 0.7329 0.5365 <.0001* 0.7284 0.5382 0.6476

Cooked Vegetable <.0001* 0.4343 0.1679 <.0001* 0.7294 0.8692 0.1098

Herbal <.0001* 0.8033 0.0014* <.0001* 0.7173 0.5514 0.9213

Black Pepper <.0001* 0.6140 0.0460 <.0001* 0.8673 0.6236 0.5396

Earthy  <.0001* 0.3959 0.2692 0.0650 0.4716 0.1957 0.2005

Vanilla <.0001* 0.2634 0.7852 <.0001* 0.3357 0.8726 0.8117

Cardboard <.0001* 0.5441 0.3529 0.0174* 0.6875 0.9854 0.5744

Chemical <.0001* 0.7088 0.4547 <.0001* 0.5527 0.8390 0.3330

Sour <.0001* 0.4467 0.0615 <.0001* 0.8074 0.9534 0.5782

Bitter <.0001* 0.5919 0.0708 0.0151* 0.8770 0.4972 0.9766

Astringent <.0001* 0.001* 0.0261* 0.0007* 0.1893 0.9834 0.9994

Viscous <.0001* 0.7148 0.9123 0.0005* 0.4474 0.7235 0.4997

Alcohol/Hot <.0001* 0.2759 0.0647 <.0001* 0.6896 0.2805 0.8640

5/17/2019

10

• 2010 vintage• 45% Lodi Muscat• 55% UC Davis Chard

• Push with nitrogen (no pump)• Filter right into bottling line

Merlot

Control – No Filtration

Filtration Through Pad 

Filter

Filtration Through Pad and 1 µm Depth Filter

Filtration Through Pad, 1 µm Depth, and 0.45 µm PVDF Filter

Time Points:1Weeks2Week4Weeks6Weeks12Weeks16 Weeks…

Rep

 1

Rep

 1

Rep

 2

Rep

 3

Rep

 1

Rep

 2

Rep

 3

Rep

 1

Rep

 2

Rep

 3

Filtration Through Pad, 1 µm Depth, and 0.45 µm PES 

Filter

Rep

 1

Rep

 2

Rep

 3

Chemical AnalysisSensory Analysis

White wine blend

5/17/2019

11

Judge Filter Bottling Day

Time  Filter* Bottling 

Filter* Time

Filter* Judge

Alcohol <.0001* 0.9281 0.6680 <.0001* 0.2914 0.7051 0.8151

Apple /Pear <.0001* 0.8379 0.3844 0.0185* 0.6743 0.7998 0.5178

Banana <.0001* 0.5204 0.4165 0.0247* 0.8011 0.2947 0.0854

Bubble Gum <.0001* 0.3357 0.7694 0.0103* 0.1635 0.6831 0.7687

Citrus <.0001* 0.1458 0.0002* <.0001* 0.2535 0.2696 0.6811

Floral <.0001* 0.4848 0.2129 <.0001* 0.7336 0.0732 0.3539

Guava <.0001* 0.8170 0.9201 <.0001* 0.5395 0.6229 0.6856

Lychee <.0001* 0.4143 0.4356 <.0001* 0.3327 0.4393 0.0612

Melon <.0001* 0.7889 0.0191* <.0001* 0.8644 0.2907 0.9791

Papaya <.0001* 0.7023 0.6911 <.0001* 0.8341 0.2450 0.9005

Peach <.0001* 0.7229 0.1058 0.0024* 0.7257 0.1945 0.9933

Pineapple <.0001* 0.7591 0.5053 <.0001* 0.6214 0.9805 0.9548

Sour <.0001* 0.9555 0.1049 <.0001* 0.8303 0.7627 0.5328

Bitter <.0001* 0.6142 0.0007* <.0001* 0.9886 0.1215 0.6047

Hot <.0001* 0.2902 0.4499 <.0001* 0.8626 0.2262 0.6796

Viscos <.0001* 0.8110 0.9943 <.0001* 0.7113 0.4634 0.7985

5/17/2019

12

5/17/2019

13

Membrane

PR

PP

Self-Cleaning from Tangential Sweeping Action

PR>PP – driving force

ANOVA Table

5/17/2019

14

Product Rep Judge Judge:Product Rep:ProductMxBerryA 4.8913* 0.5252 15.305* 0.837 2.098StonFrtA 3.6492* 0.0954 16.858* 1.0924 0.8305OakA 3.6091* 0.0686 17.005* 0.7674 1.0691caramelA 1.1663 5.4851* 14.821* 2.04* 0.1392SpiceA 1.3948 0.0152 16.787* 1.0726 1.1579CoffeeA 0.2377 0.0084 7.2885* 1.0295 0.8882EarthyA 5.4653* 2.6129 3.8949* 0.9571 1.3109SmokeA 6.4015* 0.0075 12.199* 1.3254 0.4558GrassyA 3.6153* 1.2211 6.8198* 1.3637 1.0827AlcoholA 1.1693 0.251 13.259* 0.929 0.2196SweetT 0.8626 0.0937 18.333* 0.6451 0.3407SourT 0.289 0.338 14.11* 0.3666 7.8134*BitterT 1.5156 0.0927 18.557* 1.4123 1.0004AstrinM 0.2002 4.1549 17.886* 0.6072 0.933HotM 0.8671 9.00E-04 11.923* 1.0319 1.6472ViscousM 0.7371 0.0122 19.126* 0.6917 1.0538

Filter Red Sensory

ANOVA Table

5/17/2019

15

0.00#

0.50#

1.00#

1.50#

2.00#

2.50#

3.00#

3.50#

4.00#

MxBerryA#(LSD=0.42)#

StonFrtA#(LSD=0.31)#

OakA#(LSD=0.32)#

caramelA#

SpiceA#

CoffeeA#

EarthyA#(LSD=0.35)#

SmokeA#(LSD=0.24)#

GrassyA#(LSD=0.22)#

AlcoholA#

SweetT#

SourT#

BiKerT#

AstrinM#

HotM#

ViscousM#

Overall'Means'Spiderplot'

RF1#

RF2#

RF3#

RFC#

Filter Red Sensory

+

++++

+++

+

-1.5 -0.5 0.0 0.5 1.0 1.5

-1.5

-0.5

0.5

1.5

RF1

CV1 75.2%

CV

2 1

6.7

8%

1

2345

678

9 ++++

++

+

++

-1.5 -0.5 0.0 0.5 1.0 1.5

-1.5

-0.5

0.5

1.5

RF2

CV1 75.2%

CV

2 1

6.7

8%

1234

56

78

9

++ +

++

++

++

-1.5 -0.5 0.0 0.5 1.0 1.5

-1.5

-0.5

0.5

1.5

RF3

CV1 75.2%

CV

2 1

6.7

8%

123

45

67

89 +

+

++

++

+

++

-1.5 -0.5 0.0 0.5 1.0 1.5

-1.5

-0.5

0.5

1.5

RFC

CV1 75.2%

CV

2 1

6.7

8%

12

34

56

789

CVA Over Time

-0.6 -0.4 -0.2 0.0 0.2 0.4 0.6

-0.2

0.0

0.2

0.4

Loadings Plot

Can 1 75.2%

Can

2 1

6.8

%

MxBerryA

StonFrtA

OakA

caramelA

SpiceA

CoffeeA

EarthyA

SmokeA

GrassyA

AlcoholA

SweetT

SourT

BitterT

AstrinM

HotM

ViscousM

Filter Red Sensory

5/17/2019

16

+

++

+

-1.5 -0.5 0.5 1.5

-1.5

-0.5

0.5

1.5

Week 0

CV 1 75.2%

CV

2 1

6.7

8%

RF1

RF2

RF3RFC

++

++

-1.5 -0.5 0.5 1.5

-1.5

-0.5

0.5

1.5

Week 1

CV 1 75.2%

CV

2 1

6.7

8%

RF1

RF2RF3

RFC++

+

+

-1.5 -0.5 0.5 1.5

-1.5

-0.5

0.5

1.5

Week 2

CV 1 75.2%

CV

2 1

6.7

8%

RF1RF2RF3

RFC

++++

-1.5 -0.5 0.5 1.5

-1.5

-0.5

0.5

1.5

Week 3

CV 1 75.2%

CV

2 1

6.7

8%

RF1RF2RF3RFC

+++

+

-1.5 -0.5 0.5 1.5-1

.5-0

.50

.51.

5

Week 4

CV 1 75.2%

CV

2 1

6.7

8%

RF1RF2RF3

RFC+++

+

-1.5 -0.5 0.5 1.5

-1.5

-0.5

0.5

1.5

Month 2

CV 1 75.2%

CV

2 1

6.7

8%

RF1

RF2RF3

RFC

+

+++

-1.5 -0.5 0.5 1.5

-1.5

-0.5

0.5

1.5

Month 4

CV 1 75.2%

CV

2 1

6.7

8%

RF1

RF2

RF3RFC

++ ++

-1.5 -0.5 0.5 1.5

-1.5

-0.5

0.5

1.5

Month 6

CV 1 75.2%

CV

2 1

6.7

8%

RF1RF2RF3RFC

++++

-1.5 -0.5 0.5 1.5-1

.5-0

.50

.51

.5

Month 8

CV 1 75.2%

CV

2 1

6.7

8%

RF1RF2RF3RFC

CVA Over Time

Red and white wines from cross‐flow experiment

Three pumps Centrifugal

Flexible Impeller

Progressive Cavity Two times (4.5 min and 44 min) Corresponds to 2.5 and 24 times through the pump

Two replicates of each

5/17/2019

17

5/17/2019

18

Very few differences noted during filtration for color and tannin‐no transient is obvious

No major differences observed between filtered and unfiltered wines studied

Cross flow filters seem to stabilize sensory characteristics

Pumps don’t seem to have much effect on sensory or chemical properties of wine

Chik Brenneman Paul Green Jennifer Heelan Ron Runnebaum Tarit Nimmanwudipong Sensory Panelists Rodger Pachelbel

Gallo Sonoma Silverado Vineyards

American Vineyard Foundation