1. MODELLING OF A LIGNOCELLULOSIC BIOMASS FRACTIONATION PROCESS IN A LAB-SCALE BIOREFINERY WITH HOT PRESSURIZED WATER
Álvaro Cabeza Sánchez
Cristian M. Piqueras
Francisco Sobrón Grañon
Juan García Serna
The whole work can be found in doi: /j.biortech
FPU2013/01516
CTQ R (MINECO/FEDER)

2. OBJECTIVES Biomass fractionation modelling Reactor inside simulation
Set of physical phenomena
Molecular weight
Auto-hydrolysis
Deacetylation
Reactor inside simulation
Solid phase
Liquid phase
Oligomer cleaving

3. Water+ Dissolved biomass
EXPERIMENTAL
Extraction column
Isothermal conditions (temperature ranges)
1st : 140 ºC-180 ºC (HC extraction)
2nd : 240 ºC-260 ºC (C extraction)
4 flow rates
6.5, 11.0, 15.0 and 28.0 mL/min
8 Experiments
Water+ Dissolved biomass
(TOC, C6, C5, AcetH)
Water
Biomass
(Holm oak)
(T, P, Q)

4. MODELLING: Reaction pathway
Hemicellulose
Cleaving
+ Deacetylation
+ Solubilisation
Solid
Liquid
Auto-hydrolysis

5. MODELLING: Reaction pathway
Cellulose
Cleaving + Solubilisation+ Auto-hydrolysis
Solid
Liquid

6. MODELLING: Reaction pathway
Lignin
As an inert
Structural effects
2 Hemicellulose & Cellulose
Easy to extract
Difficult to extract
Oligomer cleaving
60 members population
245 compounds

7. MODELLING: Kinetics Solid phase Based on an Auto catalytic expression
− 𝑘 𝑑 𝑗 ·𝐶 𝑆 𝑗 · 𝐹 𝑎𝑢𝑡𝑜 · 𝑖=1 𝑁 𝛼 𝑖,𝑗
− 𝑘 𝑑 𝑗 ·𝐶 𝑆 𝑗 · 𝐹 𝑎𝑢𝑡𝑜 · 𝑖=1 𝑁 𝛼 𝑖,𝑗
+ 𝑖=1 𝑁 𝛼 𝑗,𝑖 · 𝐹 𝑎𝑢𝑡𝑜 · 𝑘 𝑑 𝑖 ·𝐶 𝑆 𝑖
+ 𝑖=1 𝑁 𝛼 𝑗,𝑖 · 𝐹 𝑎𝑢𝑡𝑜 · 𝑘 𝑑 𝑖 ·𝐶 𝑆 𝑖
𝑟 𝑗 =
All the compounds produced by “j”
All the compounds that produce “j”
𝐹 𝑎𝑢𝑡𝑜 = 1−0.99·𝑓 𝑚 𝑗 , 𝑚 𝑇 −1 𝛽 𝑖,𝑗
𝛼 𝑖,𝑗 =𝑓 𝑀 𝑊 𝑖 , 𝑀 𝑊 𝑗

8. MODELLING: Kinetics Liquid phase Similar to solid phase
Auto-hydrolysis: proton order as a temperature
function
𝑟 𝑗 =− 𝑘 𝐿 𝑗 ·𝐶 𝐿 𝑗 · 𝐶 𝐻 + 𝑛 1 · 𝑖=1 𝑁 𝛼 𝑖,𝑗 + 𝑖=1 𝑁 𝛼 𝑗,𝑖 · 𝐶 𝐻 + 𝑛 1 · 𝑘 𝐿 𝑖 ·𝐶 𝐿 𝑖
𝐾= 𝑒 𝐴− 𝐸 1 +𝑓( 𝐶 𝐻 𝑅·𝑇 = 𝑒 𝐴 · 𝑒 − 𝐸 1 𝑅·𝑇 · 𝑒 − 𝑓 𝐶 𝐻 + 𝑅𝑇
𝐾= 𝑒 𝐴 · 𝑒 − 𝐸 1 𝑅·𝑇 · 𝐶 𝐻 + − 1 𝑅·𝑇
𝑓 𝐶 𝐻 + = ln 𝐶 𝐻 +
𝑛 1 =− 1 𝑅𝑇
𝑛 1 =𝑓(𝑇)

9. MODELLING: Mass balances
Liquid phase
Solid phase
𝝏𝑪 𝑳 𝒋 𝝏𝒕 = 1 ℰ · 𝒓 𝒋 − 𝒖 𝑳 · 𝝏𝑪 𝑳 𝒋 𝝏𝒛 −𝝋· 𝑪 𝑳 𝒋 · 𝝏 𝑪 𝒕 𝝏𝒕 + 𝒌 𝒋 ·𝒂· 𝑪 𝑳 𝒋 ∗ − 𝑪 𝑳 𝒋
Accumulation = reaction + convection + porosity changes effects + mass transfer
𝝏𝑪 𝑺 𝒋 𝝏𝒕 = 1 1−ℰ · 𝒓 𝒋 −𝝋· 𝑪 𝑺 𝒋 · 𝝏 𝑪 𝒕 𝝏𝒕 − 𝒌 𝒋 ·𝒂· 𝑪 𝑳 𝒋 ∗ − 𝑪 𝑳 𝒋
𝜕 1−ℰ · 𝐶 𝑡 − 𝑗=1 𝑗=𝑁 𝐶 𝑆 𝑗 𝜕𝑡 =0
ℰ=1−𝜑· 𝐶 𝑡
Inert mass balance
Porosity as function of solid concentration

10. RESULTS: Fittings TOC Q=6.5 mL/min & T=180-260 ºC A.A.D=30.15%
(Average A.A.D=42.62%)
(Average A.A.D=19.06%)

11. RESULTS: Product simulation
AcetH C5
(Average A.A.D=34.13%)
A.A.D=30.36%
(Average A.A.D=57.41%)
A.A.D=30.30% C6
A.A.D=32.36%
(Average A.A.D=42.64%)

12. RESULTS: Solid cleaving simulation
Hemicellulose
Cellulose

13. CONCLUSIONS Modelling for biomass fractionation Simulation
Subcritical range (140 ºC – 260 ºC)
Flow & temperature effect
Set of physical phenomena
MW effect
Low TOC deviation (<20%)
General for lignocellulosic biomass
Simulation
Oligomer cleaving
Reactor behaviour

14. THANK FOR YOUR ATTENTION
Álvaro Cabeza Sánchez
Cristian M. Piqueras
Francisco Sobrón Grañon
Juan García Serna
The whole work can be found in doi: /j.biortech
FPU2013/01516
CTQ R (MINECO/FEDER)

15. AAD considerations Reasons Solution
TOC vs HPLC data  discrepancy around 20%
Dilution  fluctuating profiles
Sample biodiversity  different behaviour at the same conditions
Solution
Focusing the analysis on HPLC data
Optimal conditions for hemicelluloses extraction from Eucalyptus globulus wood: hydrothermal treatment in a semi-continuous reactor, Fuel Processing Technology (2016) 350–360

16. AAD considerations pH C6 C5 A.A.D=2.91% A.A.D=14.37%
(Average A.A.D=5.57%)
A.A.D=2.91%
(Average A.A.D=13.17%)
A.A.D=14.37% C5
(Average A.A.D=10.49%)
A.A.D=16.27%