Recycling of Meat and Bone Meal Animal F
Recycling of Meat and Bone Meal
Animal Feed by Vacuum Pyrolysis
A . C H A A L A A N D C . R O Y*
De´partem en t de Ge´n ie Ch im iqu e, Un iversite´ Laval,
Sain te-Foy, Qu e´bec, Can ada G1K 7P4
Due to the recent bovine spongiform encephalopathy
(BSE) crisis in the European beef industry, the use of animalderived products to feed cattle is now severely restricted.
Large quantities of w aste animal meat and bone meal
(M BM ), also know n as animal flour, have to be safely
disposed of or transformed. One disposal option is pyrolysis.
Vacuum pyrolysis of an animal flour sample has been
performed in a laboratory reactor. The results obtained
revealed that vacuum pyrolysis can be an attractive alternative
to incineration and cement kilns. The process generated
a combustible gas (15.1 w t %), a high calorific value oil (35.1
w t %), a solid residue rich in minerals (39.1 w t %), and
an aqueous phase rich in organics (10.7 w t %). The gas and
the aqueous phase can be used to provide heat to the
vacuum pyrolysis reactor and the M BM drying unit. The
oil can be used alone or mixed w ith petroleum products as
a fuel in boilers or gas turbines. Conversion of animal
w aste by pyrolysis into fuels can contribute to the reduction
of greenhouse gases. It is suggested to use the solid
residue for agricultural soil enrichment in minerals and as
a soil moisturizer.
Introduction
Th e m ad cow disease h as h ad an n oyin g rep ercu ssion s in th e
an im al feed in du stry. In fact, an im al flou r an d byp rodu cts
th erefrom , wh ich are n ow p rodu ced in h u ge qu an tities, h ave
to be safely disp osed of. Th ese p rodu cts m u st be elim in ated
or safely recycled. In Fran ce alon e, the total am oun t of an im al
residu es wh ich n eed to be elim in ated is greater th an 3 Mton s
p er year (1). Th is in clu des abou t 850000 ton s of defated flou r
an d 300000 ton s of fatty p rodu cts. Existin g tech n ologies can
treat on ly 350000 ton s of flou r an d th e derived fatty
byp rodu cts.
To avoid h u ge storage costs of an im al flou r an d cru de
an im al residu es, n ew solu tion s h ave to be in vestigated.
In cin eration is a p oten tial solu tion , bu t th e flu ffy n atu re of
th e flou r is ch allen gin g as p articles fin d th eir way in th e
em ission s an d th e ash residu e m elts an d stick in side th e
in cin eration ch am ber. It h as been rep orted by McDon n ell
et al. (2) th at, to effectively bu rn m eat an d bon e m eal, it is
n ecessary to m ix th em with m illed p eat at a con cen tration
of 35 wt %. A qu ick bu t tran sitory solu tion au th orized by th e
Fren ch govern m en t for th e elim in ation of an im al flou r is in
th e cem en t in du stry as a solid fu el (1).
Sin ce th e cem en t in du stry in Fran ce is ren own ed for its
su itability to u se diverse typ es of fu el residu es su ch as u sed
lu brican t oils, solven ts, p ain ts, u sed tires, an d n ow an im al
flou r, it h as th e p oten tial to bu rn 450000 ton s/ year of an im al
flou r after 2002. Cu rren tly on ly 16 ou t of 33 cem en t p lan ts
in Fran ce con su m e flou rs (1). Accordin g to th e MIEFA
(Mission in term in iste´ rielle p ou r l’e´ lim in ation des farin es
an im ales), seven oth er cem en t factories wh ich cou ld con su m e 80000 ton s/ year of an im al flou r are requ ired (3, 4).
Som e in vestigation s h ave been p erform ed on th e p ossibility
of m akin g p h en olic briqu ettes with a m ixtu re of con crete
an d an im al flou r. Th is idea cou ld n ot be realized at a large
scale du e to th e h igh sen sitivity of th e briqu ettes to water.
The briquettes decom pose with tim e an d can gen erate phen ol
derivative com p ou n ds (1).
An im al flou r con tain s a large am ou n t of fats an d oth er
organ ic com p ou n ds. It can be con verted by th erm al degradation p rocesses su ch as p yrolysis in to tran sp ortable,
storable, an d workable fu els. Vacu u m p yrolysis h as p roven
to be ap p licable for th e u p gradin g of a large variety of solid
an d sem iliqu id organ ic wastes (5-11). Th e p rocess is
com m ercialized by Pyrovac In tern ation al In c. u n der th e
tradem ark of Pyrocyclin g (7). Th e objective of th is p ap er is
to dem on strate th e p oten tial of th e vacu u m p yrolysis
tech n ology (Pyrocyclin g) for th e treatm en t of an im al flou r.
Th e resu lts obtain ed at th e laboratory scale an d h erein
rep orted p rovide th e backgrou n d data for a tech n ological
develop m en t of th is solu tion at th e in du strial scale.
Experimental Section
Materials. Th e m eat an d bon e m eal (MBM) feedstock h erein
called an im al flou r was p rovided by Alex Cou tu re In c.
(Ch arn y, Qu e´ bec, Can ada). Th e m aterial was received as a
p owder an d h ad to be p elletized as its flu ffy n atu re did n ot
en able ap p lication of vacu u m su ction in to th e reactor. No
bin der was added du rin g th e extru sion p rocess. Th e p ellets
h ad an average m oistu re con ten t of 3.5 wt %. Th e size of th e
p ellets varied between 5 an d 20 m m with a diam eter of 5
m m . Th e p ellets were h ard en ou gh for treatm en t in a batch
p rocess. Th ey wou ld n eed to be fu rth er h arden ed h owever
if p rocessed in a con tin u ou s feed reactor.
Analysis. Proxim ate an alyses of th e feedstock an d th e
solid residu e obtain ed were determ in ed by m ean s of a LECO
TGA-601 an alyzer, wh ereas th e elem en tal com p osition
(carbon , h ydrogen , an d n itrogen ) was determ in ed in a LECO
CHN-2000 elem en tal an alyzer. Ph ysicoch em ical ch aracteristics of th e oils were determ in ed accordin g to ap p rop riate
ASTM m eth ods. Th e gross calorific valu e was determ in ed
accordin g to ASTM D-240 for solids an d ASTM D-4809 for
liqu ids.
A SSC/ 5200 TG/ DTG (220) m icrobalan ce from Seiko was
u sed for th e th erm ogravim etric tests. Sm all p ellet sam p les
weigh in g 7-8 m g an d 1-2 m m in dim en sion s were h eated
from room tem p eratu re to 550 °C, u n der a n itrogen flow of
150 m L/ m in at a h eatin g rate of 10 °C/ m in . Th e con dition s
for th e th erm ogravim etric an alyses were ch osen n ear th e
con dition s u sed in th e p ilot p lan t for th e p yrolysis of sim ilar
residu es su ch as u rban sewage slu dge, softwood bark p ellets,
an d bagasse.
Th e m etal con ten t of th e solid m aterials was m easu red
by in ductively coupled plasm a atom ic em ission spectroscopy
(ICP-AES) on a rep resen tative sam p le of ash dissolved in
m in eral acids. Organ ic liqu ids were an alyzed by gas ch rom atography-m ass spectrom etry (GC-MS) usin g an HP-5890
gas ch rom atograp h equ ip p ed with a 30 m × 0.25 m m i.d. J
& W fu sed silica cap illary colu m n DB5 cou p led to an HP5970 m ass-selective detector.
Pyrolysis Test. Aseries of laboratory tests were p erform ed
p rior to fixin g th e best p yrolysis con dition s. Th e feedstock
as received was in a p owder form . Its sp ecific gravity was
very low. Du rin g th e p yrolysis test, in th e exten sive degrada-
FIGURE 1. Vacuum pyrolysis laboratory-scale installation.
FIGURE 2. TG and DTG of the animal flour sample.
tion p rocess, a large am ou n t of p owder left th e reactor with
th e vap ors gen erated, cau sin g th e form ation of p lu gs in th e
lin es of th e con den sin g system . To avoid th is p h en om en on ,
th e feedstock was p elletized.
Figu re 1 illu strates th e p rocess sch em atics. Th e reactor
u sed for th e p yrolysis ru n s h as a workin g volu m e of 1 L. Th e
stain less steel reactor was in stalled in an electrically h eated
tu bu lar fu rn ace. Th e gas an d vap ors p rodu ced du rin g th e
p yrolysis p assed th rou gh a stain less steel trap an d fou r cold
trap s con n ected in series. Th e stain less steel trap was
m ain tain ed at room tem p eratu re. Th e last fou r trap s m ade
of Pyrex were m ain tain ed at -15, -40, -78, an d -78 °C. Th e
n on con den sable gas was p u m p ed in to a stain less steel
con tain er, p reviou sly set u n der vacu u m by m ean s of a
vacu u m p u m p .
The test tem perature was m easured usin g a therm ocoup le
in stalled in th e feedstock bed in side th e reactor. Both th e
p ressu re an d th e tem p eratu re were m easu red an d recorded
every 30 s, by m ean s of a data acqu isition system .
Wh en th e p ressu re of th e recovered gases rem ain ed
con stan t for 30 m in , th e p yrolysis of th e feedstock was
con sidered to be com p lete. Th e h eatin g of th e reactor was
th en term in ated, an d n itrogen was in trodu ced to avoid an y
oxidation reaction s. On ce th e reactor an d th e rest of th e
ap p aratu s h ad cooled to room tem p eratu re, th e p yrolysis
p rodu cts were recovered an d weigh ed.
Asam ple of 516.5 g of pelletized an im al flour was pyrolyzed
at a tem p eratu re of 500 °C, a total p ressu re of ap p roxim ately
4 kPa, an d a h eatin g rate of 15 °C/ m in . Th e batch p yrolysis
test lasted 170 m in . Th e en d p yrolysis tem p eratu re was
selected on th e basis of a series of TG tests (Figu re 2).
Results and Discussion
Characterization of the Feedstock. Th e p roxim ate an alysis
p resen ted in Table 1 sh ows th at th e feedstock con tain s a
h igh am ou n t of volatile m atters. Th e h igh ash con ten t
(24.8 wt %) is attribu ted to th e p resen ce of bon es in th e
feedstock, an d as a resu lt th is residu e h as a low calorific
valu e (20 MJ/ kg). Th e low fixed carbon (8.0 wt %) an d th e
h igh oxygen (17.7 wt %) an d n itrogen (8.9 wt %) con ten ts are
also n oticeable.
Therm ogravim etric Analysis of the Feedstock. Figu re 2
p resen ts th e TG (th erm ogravim etric) an d DTG (differen tial
th erm ogravim etric) cu rves of th e an im al flou r sam p le. Th ree
sh ou lders an d on e p eak are observed on th e DTG cu rve. Th e
m ass loss (3.25 wt %) rep resen ted by th e first sh ou lder in th e
tem p eratu re ran ge of 50-148 °C is attribu ted to th e deh ydration of th e an im al flou r sam p le. Th e secon d sh ou lder,
wh ich is observed between 148 an d 225 °C, m ay be du e to
th e evap oration of low m olecu lar weigh t com p ou n ds con tain ed in the an im al flour or/ an d the decom position reaction s
TABLE 1. Physicochemical Properties of the Feedstock and
the Solid Residue
properties
m oisture content, w t %
proxim ate analysis (anhydrous basis, w t %)
volatile m atters
ash
fixed carbon
elem ental analysis (anhydrous basis, w t %)
carbon
hydrogen
nitrogen
oxygen a
gross calorific value, M J/kg
a
feedstock
solid
residue
3.5
0.0
67.2
24.8
8.0
21.5
56.2
22.3
42.8
5.8
8.9
17.7
20
31.8
1.4
5.2
5.4
11.5
Determ ined by difference.
TABLE 2. Metal Content in the Feedstock and the Solid
Residue (mg/kg) a
element
feedstock
solid
residue
element
feedstock
solid
residue
Ca
P
Na
K
Mg
Fe
Zn
Mn
Cu
Cr
67207
31842
8716
6339
1599
560.5
99.8
35.2
22.0
5.9
141167
72363
17872
13523
3381
883.8
183.9
74.1
41.5
5.9
Ti
Ni
Co
Pb
As
Se
Mo
Al
V
Cd
3.5
2.9
0.4
Animal Feed by Vacuum Pyrolysis
A . C H A A L A A N D C . R O Y*
De´partem en t de Ge´n ie Ch im iqu e, Un iversite´ Laval,
Sain te-Foy, Qu e´bec, Can ada G1K 7P4
Due to the recent bovine spongiform encephalopathy
(BSE) crisis in the European beef industry, the use of animalderived products to feed cattle is now severely restricted.
Large quantities of w aste animal meat and bone meal
(M BM ), also know n as animal flour, have to be safely
disposed of or transformed. One disposal option is pyrolysis.
Vacuum pyrolysis of an animal flour sample has been
performed in a laboratory reactor. The results obtained
revealed that vacuum pyrolysis can be an attractive alternative
to incineration and cement kilns. The process generated
a combustible gas (15.1 w t %), a high calorific value oil (35.1
w t %), a solid residue rich in minerals (39.1 w t %), and
an aqueous phase rich in organics (10.7 w t %). The gas and
the aqueous phase can be used to provide heat to the
vacuum pyrolysis reactor and the M BM drying unit. The
oil can be used alone or mixed w ith petroleum products as
a fuel in boilers or gas turbines. Conversion of animal
w aste by pyrolysis into fuels can contribute to the reduction
of greenhouse gases. It is suggested to use the solid
residue for agricultural soil enrichment in minerals and as
a soil moisturizer.
Introduction
Th e m ad cow disease h as h ad an n oyin g rep ercu ssion s in th e
an im al feed in du stry. In fact, an im al flou r an d byp rodu cts
th erefrom , wh ich are n ow p rodu ced in h u ge qu an tities, h ave
to be safely disp osed of. Th ese p rodu cts m u st be elim in ated
or safely recycled. In Fran ce alon e, the total am oun t of an im al
residu es wh ich n eed to be elim in ated is greater th an 3 Mton s
p er year (1). Th is in clu des abou t 850000 ton s of defated flou r
an d 300000 ton s of fatty p rodu cts. Existin g tech n ologies can
treat on ly 350000 ton s of flou r an d th e derived fatty
byp rodu cts.
To avoid h u ge storage costs of an im al flou r an d cru de
an im al residu es, n ew solu tion s h ave to be in vestigated.
In cin eration is a p oten tial solu tion , bu t th e flu ffy n atu re of
th e flou r is ch allen gin g as p articles fin d th eir way in th e
em ission s an d th e ash residu e m elts an d stick in side th e
in cin eration ch am ber. It h as been rep orted by McDon n ell
et al. (2) th at, to effectively bu rn m eat an d bon e m eal, it is
n ecessary to m ix th em with m illed p eat at a con cen tration
of 35 wt %. A qu ick bu t tran sitory solu tion au th orized by th e
Fren ch govern m en t for th e elim in ation of an im al flou r is in
th e cem en t in du stry as a solid fu el (1).
Sin ce th e cem en t in du stry in Fran ce is ren own ed for its
su itability to u se diverse typ es of fu el residu es su ch as u sed
lu brican t oils, solven ts, p ain ts, u sed tires, an d n ow an im al
flou r, it h as th e p oten tial to bu rn 450000 ton s/ year of an im al
flou r after 2002. Cu rren tly on ly 16 ou t of 33 cem en t p lan ts
in Fran ce con su m e flou rs (1). Accordin g to th e MIEFA
(Mission in term in iste´ rielle p ou r l’e´ lim in ation des farin es
an im ales), seven oth er cem en t factories wh ich cou ld con su m e 80000 ton s/ year of an im al flou r are requ ired (3, 4).
Som e in vestigation s h ave been p erform ed on th e p ossibility
of m akin g p h en olic briqu ettes with a m ixtu re of con crete
an d an im al flou r. Th is idea cou ld n ot be realized at a large
scale du e to th e h igh sen sitivity of th e briqu ettes to water.
The briquettes decom pose with tim e an d can gen erate phen ol
derivative com p ou n ds (1).
An im al flou r con tain s a large am ou n t of fats an d oth er
organ ic com p ou n ds. It can be con verted by th erm al degradation p rocesses su ch as p yrolysis in to tran sp ortable,
storable, an d workable fu els. Vacu u m p yrolysis h as p roven
to be ap p licable for th e u p gradin g of a large variety of solid
an d sem iliqu id organ ic wastes (5-11). Th e p rocess is
com m ercialized by Pyrovac In tern ation al In c. u n der th e
tradem ark of Pyrocyclin g (7). Th e objective of th is p ap er is
to dem on strate th e p oten tial of th e vacu u m p yrolysis
tech n ology (Pyrocyclin g) for th e treatm en t of an im al flou r.
Th e resu lts obtain ed at th e laboratory scale an d h erein
rep orted p rovide th e backgrou n d data for a tech n ological
develop m en t of th is solu tion at th e in du strial scale.
Experimental Section
Materials. Th e m eat an d bon e m eal (MBM) feedstock h erein
called an im al flou r was p rovided by Alex Cou tu re In c.
(Ch arn y, Qu e´ bec, Can ada). Th e m aterial was received as a
p owder an d h ad to be p elletized as its flu ffy n atu re did n ot
en able ap p lication of vacu u m su ction in to th e reactor. No
bin der was added du rin g th e extru sion p rocess. Th e p ellets
h ad an average m oistu re con ten t of 3.5 wt %. Th e size of th e
p ellets varied between 5 an d 20 m m with a diam eter of 5
m m . Th e p ellets were h ard en ou gh for treatm en t in a batch
p rocess. Th ey wou ld n eed to be fu rth er h arden ed h owever
if p rocessed in a con tin u ou s feed reactor.
Analysis. Proxim ate an alyses of th e feedstock an d th e
solid residu e obtain ed were determ in ed by m ean s of a LECO
TGA-601 an alyzer, wh ereas th e elem en tal com p osition
(carbon , h ydrogen , an d n itrogen ) was determ in ed in a LECO
CHN-2000 elem en tal an alyzer. Ph ysicoch em ical ch aracteristics of th e oils were determ in ed accordin g to ap p rop riate
ASTM m eth ods. Th e gross calorific valu e was determ in ed
accordin g to ASTM D-240 for solids an d ASTM D-4809 for
liqu ids.
A SSC/ 5200 TG/ DTG (220) m icrobalan ce from Seiko was
u sed for th e th erm ogravim etric tests. Sm all p ellet sam p les
weigh in g 7-8 m g an d 1-2 m m in dim en sion s were h eated
from room tem p eratu re to 550 °C, u n der a n itrogen flow of
150 m L/ m in at a h eatin g rate of 10 °C/ m in . Th e con dition s
for th e th erm ogravim etric an alyses were ch osen n ear th e
con dition s u sed in th e p ilot p lan t for th e p yrolysis of sim ilar
residu es su ch as u rban sewage slu dge, softwood bark p ellets,
an d bagasse.
Th e m etal con ten t of th e solid m aterials was m easu red
by in ductively coupled plasm a atom ic em ission spectroscopy
(ICP-AES) on a rep resen tative sam p le of ash dissolved in
m in eral acids. Organ ic liqu ids were an alyzed by gas ch rom atography-m ass spectrom etry (GC-MS) usin g an HP-5890
gas ch rom atograp h equ ip p ed with a 30 m × 0.25 m m i.d. J
& W fu sed silica cap illary colu m n DB5 cou p led to an HP5970 m ass-selective detector.
Pyrolysis Test. Aseries of laboratory tests were p erform ed
p rior to fixin g th e best p yrolysis con dition s. Th e feedstock
as received was in a p owder form . Its sp ecific gravity was
very low. Du rin g th e p yrolysis test, in th e exten sive degrada-
FIGURE 1. Vacuum pyrolysis laboratory-scale installation.
FIGURE 2. TG and DTG of the animal flour sample.
tion p rocess, a large am ou n t of p owder left th e reactor with
th e vap ors gen erated, cau sin g th e form ation of p lu gs in th e
lin es of th e con den sin g system . To avoid th is p h en om en on ,
th e feedstock was p elletized.
Figu re 1 illu strates th e p rocess sch em atics. Th e reactor
u sed for th e p yrolysis ru n s h as a workin g volu m e of 1 L. Th e
stain less steel reactor was in stalled in an electrically h eated
tu bu lar fu rn ace. Th e gas an d vap ors p rodu ced du rin g th e
p yrolysis p assed th rou gh a stain less steel trap an d fou r cold
trap s con n ected in series. Th e stain less steel trap was
m ain tain ed at room tem p eratu re. Th e last fou r trap s m ade
of Pyrex were m ain tain ed at -15, -40, -78, an d -78 °C. Th e
n on con den sable gas was p u m p ed in to a stain less steel
con tain er, p reviou sly set u n der vacu u m by m ean s of a
vacu u m p u m p .
The test tem perature was m easured usin g a therm ocoup le
in stalled in th e feedstock bed in side th e reactor. Both th e
p ressu re an d th e tem p eratu re were m easu red an d recorded
every 30 s, by m ean s of a data acqu isition system .
Wh en th e p ressu re of th e recovered gases rem ain ed
con stan t for 30 m in , th e p yrolysis of th e feedstock was
con sidered to be com p lete. Th e h eatin g of th e reactor was
th en term in ated, an d n itrogen was in trodu ced to avoid an y
oxidation reaction s. On ce th e reactor an d th e rest of th e
ap p aratu s h ad cooled to room tem p eratu re, th e p yrolysis
p rodu cts were recovered an d weigh ed.
Asam ple of 516.5 g of pelletized an im al flour was pyrolyzed
at a tem p eratu re of 500 °C, a total p ressu re of ap p roxim ately
4 kPa, an d a h eatin g rate of 15 °C/ m in . Th e batch p yrolysis
test lasted 170 m in . Th e en d p yrolysis tem p eratu re was
selected on th e basis of a series of TG tests (Figu re 2).
Results and Discussion
Characterization of the Feedstock. Th e p roxim ate an alysis
p resen ted in Table 1 sh ows th at th e feedstock con tain s a
h igh am ou n t of volatile m atters. Th e h igh ash con ten t
(24.8 wt %) is attribu ted to th e p resen ce of bon es in th e
feedstock, an d as a resu lt th is residu e h as a low calorific
valu e (20 MJ/ kg). Th e low fixed carbon (8.0 wt %) an d th e
h igh oxygen (17.7 wt %) an d n itrogen (8.9 wt %) con ten ts are
also n oticeable.
Therm ogravim etric Analysis of the Feedstock. Figu re 2
p resen ts th e TG (th erm ogravim etric) an d DTG (differen tial
th erm ogravim etric) cu rves of th e an im al flou r sam p le. Th ree
sh ou lders an d on e p eak are observed on th e DTG cu rve. Th e
m ass loss (3.25 wt %) rep resen ted by th e first sh ou lder in th e
tem p eratu re ran ge of 50-148 °C is attribu ted to th e deh ydration of th e an im al flou r sam p le. Th e secon d sh ou lder,
wh ich is observed between 148 an d 225 °C, m ay be du e to
th e evap oration of low m olecu lar weigh t com p ou n ds con tain ed in the an im al flour or/ an d the decom position reaction s
TABLE 1. Physicochemical Properties of the Feedstock and
the Solid Residue
properties
m oisture content, w t %
proxim ate analysis (anhydrous basis, w t %)
volatile m atters
ash
fixed carbon
elem ental analysis (anhydrous basis, w t %)
carbon
hydrogen
nitrogen
oxygen a
gross calorific value, M J/kg
a
feedstock
solid
residue
3.5
0.0
67.2
24.8
8.0
21.5
56.2
22.3
42.8
5.8
8.9
17.7
20
31.8
1.4
5.2
5.4
11.5
Determ ined by difference.
TABLE 2. Metal Content in the Feedstock and the Solid
Residue (mg/kg) a
element
feedstock
solid
residue
element
feedstock
solid
residue
Ca
P
Na
K
Mg
Fe
Zn
Mn
Cu
Cr
67207
31842
8716
6339
1599
560.5
99.8
35.2
22.0
5.9
141167
72363
17872
13523
3381
883.8
183.9
74.1
41.5
5.9
Ti
Ni
Co
Pb
As
Se
Mo
Al
V
Cd
3.5
2.9
0.4