Fuel 184 (2016) 780–791

Contents lists available at ScienceDirect

Fuel
journal homepage: www.elsevier.com/locate/fuel

Full Length Article

An experimental investigation of the combustion performance of human

faeces
Tosin Onabanjo, Athanasios J. Kolios ⇑, Kumar Patchigolla, Stuart T. Wagland, Beatriz Fidalgo, Nelia Jurado,
Dawid P. Hanak, Vasilije Manovic, Alison Parker, Ewan McAdam, Leon Williams, Sean Tyrrel, Elise Cartmell
Cranfield University, Cranfield, Bedfordshire MK43 0AL, United Kingdom

h i g h l i g h t s

Dry human faeces have a Higher Heating Value (HHV) of 24 MJ/kg.

Faeces combustion was investigated using a bench-scale downdraft combustor test rig.

Combustion temperature of 431–558 °C was achieved at air flow rate of 10–18 L/min.

Fuel burn rate of 1.5–2.3 g/min was achieved at air flow rate of 10–18 L/min.

Combustion temperature of up to 600 ± 10 °C can handle 60 wt.% moisture in faeces.

a r t i c l e i n f o a b s t r a c t

Article history: Poor sanitation is one of the major hindrances to the global sustainable development goals. The Reinvent
Received 10 June 2016 the Toilet Challenge of the Bill and Melinda Gates Foundation is set to develop affordable, next-
Received in revised form 14 July 2016 generation sanitary systems that can ensure safe treatment and wide accessibility without compromise
Accepted 19 July 2016
on sustainable use of natural resources and the environment. Energy recovery from human excreta is
Available online 27 July 2016
likely to be a cornerstone of future sustainable sanitary systems. Faeces combustion was investigated
using a bench-scale downdraft combustor test rig, alongside with wood biomass and simulant faeces.
Keywords:
Parameters such as air flow rate, fuel pellet size, bed height, and fuel ignition mode were varied to estab-
Faecal biomass
Combustion
lish the combustion operating range of the test rig and the optimum conditions for converting the faecal
Smouldering biomass to energy. The experimental results show that the dry human faeces had a higher energy content
Non-sewered sanitary systems (25 MJ/kg) than wood biomass. At equivalence ratio between 0.86 and 1.12, the combustion tempera-
Nano membrane toilet ture and fuel burn rate ranged from 431 to 558 °C and 1.53 to 2.30 g/min respectively. Preliminary results
for the simulant faeces show that a minimum combustion bed temperature of 600 ± 10 °C can handle fae-
ces up to 60 wt.% moisture at optimum air-to-fuel ratio. Further investigation is required to establish the
appropriate trade-off limits for drying and energy recovery, considering different stool types, moisture
content and drying characteristics. This is important for the design and further development of a self-
sustained energy conversion and recovery systems for the NMT and similar sanitary solutions.
Ó 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://
creativecommons.org/licenses/by/4.0/).

1. Introduction world’s population (about 2.4 billion people) lack adequate sanita-
tion facilities, particularly in developing countries [2]. In these
Human faeces is a rich source of biomass having a mixture of areas, >90% of the faeces generated are disposed into the open
undigested fats, protein, water, polysaccharide, bacterial biomass, without treatment, polluting surrounding lakes and rivers [3]. Even
gut secretions, cell shedding and ash [1]. This useful resource is in communities with modern sanitary systems, wastewater often
typically treated as a waste material, and openly disposed in the leak into the environment, due to improper usage and
environment by the nearly 1 billion people world-wide who have maintenance of septic systems, putting the groundwater at risk
no access or do not use a toilet. It is estimated that 40% of the of contamination, especially areas with high water table. These
scenarios pose a number of health and environmental hazards
⇑ Corresponding author at: School of Water, Energy, and Environment, Cranfield including the outbreak of infectious diseases and parasitic worms.
University, Cranfield, Bedfordshire MK430AL, United Kingdom. The world’s population is projected to reach close to 10 billion by
E-mail address: a.kolios@cranfield.ac.uk (A.J. Kolios). 2050 [4]; consequently, poor sanitation is projected to increase due

http://dx.doi.org/10.1016/j.fuel.2016.07.077
0016-2361/Ó 2016 The Authors. Published by Elsevier Ltd.
This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
 T. Onabanjo et al. / Fuel 184 (2016) 780–791 781

to rapid urbanization and overburdening of already stressed waste co-combustion of polyethylene waste and charcoal in a combus-
treatment systems. Therefore, the development and universal tion cell and compared the outcomes with the combustion of
access of improved sanitation facilities have become a global human faeces. Their study showed that the temperature
priority. profile from human faeces combustion was more uniform than
The Reinvent the Toilet Challenge (RTTC) of the Bill and Melinda co-combustion of polyethylene and charcoal. Earlier studies by
Gates Foundation is set to develop affordable, next-generation san- the same authors [21] focussed on the ignition properties of these
itary systems that can work without connection to external water, fuels; however, little is shown on the operating conditions for
energy or sewerage systems [5]. The Nano-Membrane toilet (NMT) handling the combustion of human faeces.
is an example of such a unique off-grid, household-scale toilet Unlike well-established fuels with uniform fuel characteristics
solution that is being developed at Cranfield University to safely such as coal, the physical and chemical characteristics of human
treat human waste onsite [6]. This unit will integrate a compact faeces vary with nutritional intake, health status, gender, body
energy conversion system that can thermally-treat human faeces, weight and age of individuals [1]. Faeces also possess complex
without external energy and water supply. The energy recovered compositional characteristics, such as the presence of a viscous
can then be used to meet the toilet’s energy requirement to ensure ‘‘sticky” substance, possibly from the linings of the intestinal wall,
the self-sustained operation of the NMT unit. which makes handling and pre-treating of the samples difficult.
Among thermochemical conversion processes, combustion is a There is therefore the need to understand the combustion pro-
promising and the most mature technology targeted for treating cesses of human faeces and establish the right operating range
biomass, [7]. It is an exothermic reaction that ensures the complete for fuel conversion, considering sample variabilities and uncertain-
conversion of a fuel in the presence of an oxidant and heat, with ties. This study describes the combustion performance of a bench-
the product gas largely constituted by carbon dioxide (CO2). Com- scale downdraft combustor test rig when utilised for human faeces
bustion can consist of rapid oxidation of the fuel, which is charac- combustion. Initial set of analyses were conducted using simulant
terised by high temperatures of >1000 °C with the visible presence faeces and wood biomass to ensure repeatability, for fuel compar-
of flames, or a slow, progressive, flameless, and relatively low tem- ison and to understand the combustion operating conditions of the
perature reaction, referred to as smouldering. The latter involves test rig. Parameters such as fuel moisture content, air flow rate,
the oxidation of the fuel in the gas phase surrounding the fuel fuel pellet size and ignition mode were investigated. Performance
[8] with progressive burn, spread and heat release rate. Peak tem- evaluation was carried out on the basis of combustion tempera-
peratures for smouldering are reported in the range of 500–700 °C ture, fuel burn rate, modified combustion efficiency (MCE) and car-
[8], which could be low as 250 °C [9]. It involves the thermal degra- bon conversion efficiency (gCCE).
dation of the fuel, evolution of volatiles, and resulting visible glow
of heat that propagates into flames, depending on oxygen availabil-
ity, the presence of stable source of heat (external or previously 2. Methods
heated material) as well as feedstock composition and characteris-
tics [10,11]. 2.1. Fuel characterization
Smouldering has recently gained importance in practice as an
applicable technology. Some of its recent applications include About 3 kg of fresh human faeces was collected and stored in a
remediation of coal tar [12]; combustion of wood biomass freezer at 85 °C, over a period of two weeks to preserve and pre-
[10,11,13]; treatment of bio-solids from wastewater treatment vent microbial degradation of the samples. The frozen samples
plant [14]; remediation of oil contaminated soil [15] and faeces were thawed at room temperature and mixed until a uniform con-
treatment [16,17]. Yermán et al. [16] showed that a self- sistency was obtained. The homogenised human faeces (HF) sam-
sustained smouldering combustion can be applied to treat moist ple was dried at 45 ± 5 °C in a GENLAB Hot Air Oven to constant
faeces under given sand pack height, sand-to-faeces ratio, air-to- weight. The limited drying temperature was applied to prevent
fuel ratio, and faecal moisture limit. Their studies examined the the loss of volatile matter. Simulant faeces (SF) were prepared
smouldering combustion of surrogate faeces and validated the pro- using the recipe outlined in Table 1 [22], while wood biomass
cess for faeces treatment using dog faeces. Wall et al. [17] investi- (WB) was sourced locally. The relative percentages of carbon (C),
gated the influence of moisture content on smouldering velocity hydrogen (H) and nitrogen (N) in the samples were determined
and upscaling of the process. Both studies showed that smoulder- using a thermal elemental analyser (Vario ELIII CHN) according
ing can be applied for faeces treatment in on-site sanitation sys- to BS EN ISO 16948. The moisture content of the sample was deter-
tems, although, this was not demonstrated with human faeces. mined at 105 ± 5 °C using the protocol outlined in BS EN 14774-3.
The operational parameters such as sand pack height and sand- The volatile matter and ash content were determined using a Car-
to-faeces mass ratio require sand as a porous medium in the com- bolite muffle furnace with the heating conditions in BS EN 15148
bustor, which are not typical characteristics of a conventional com- and BS EN 14775 respectively. The oxygen (O) content was
bustor. Thus, little is known about the fundamental combustion obtained by subtracting the wt.% percent of C, H, N and ash from
processes of human faeces. 100% while the fixed carbon content was obtained by subtracting
Other experimental studies have explored processes such as
pyrolysis [18] and hydrothermal carbonization [19] to treat human
Table 1
faeces; although, the development of these technologies is in the Recipe for simulant faeces [22].
early stages and could have added complexities and costs. Ward
Ingredients Dry weight (g/kg)
et al. [18] examined the HHV of real and synthetic faecal char at
pyrolysis temperature of 300–750 °C and showed that the faecal Baker’s yeast 72.8
Peanut oil 38.8
char obtained at pyrolysis temperature of 300 °C had a comparable
Miso paste 24.3
HHV with wood char. Afolabi et al. [19] employed microwave- Propylene glycol 24.3
assisted hydrothermal carbonization to treat human faeces at Cellulose powder 12.4
temperatures of 160–200 °C and residence time of 30–120 min Psyllium husk powder 24.3
under autogenous pressure, and recovered char and ammonia. Calcium phosphate 25.0
Watera 778.1
These studies confirm that human faeces have unique resource
a
recovery potentials. Monhol and Martins [20] investigated the Water was added based on the required moisture content.
782 T. Onabanjo et al. / Fuel 184 (2016) 780–791

the wt.% percent of moisture, ash and volatile matter from 100% faeces, as reported in [1]. Two fuel ignition modes were tested to
[23]. Samples were prepared in triplicates, and mean values are understand the best approach for igniting the faecal biomass: (a)
reported for each analysis in Table 2. The HHVs for these samples standard ignition as explained above, and (b) ‘booster’ ignition,
were determined using Parr Bomb Calorimeter [6400 Automatic considering the requirement of limited use of energy. For the boos-
Isoperibol]. ter ignition, the suction fan was operated only when the igniter
temperature reaches about 600 °C. Similar to the standard ignition
2.2. Experimental set-up procedure, once the fuel is ignited the igniter is switched off and
the primary air flow rate is set to the required value. All experi-
The schematic diagram of the experimental test rig used in this ments were carried out feeding the primary air at ambient temper-
study is shown in Fig. 1. This bench scale test rig was provided by ature and pressures. Primary air flow rate ranges from 10 to 18 L/
RTI International/Colorado State University, as part of the Nano- min depending on the experiment. The volumetric compositions of
membrane Toilet Phase II project of the RTTC with the objective CO, CO2, H2, and CH4 were measured at every second interval and
to investigate the thermochemical conversion processes of human recorded via a data logging system. The gas sampling unit consists
faeces. The test rig consists of a downdraft reactor with air supply, of three PTFE filtration and condensation units, connected in series
igniter, fuel grate, fuel feed-in and an ash collection system. The to the gas analyser that is based on thermal conductivity detection.
igniter is powered by an AC bench power supply. The reactor is a The analyser was calibrated at intervals with a standard gas once a
stainless-steel cylindrical tube with a diameter of 0.07 m and a day and at the start of experiment. At the end of the experiment,
height of 0.23 m from the fuel inlet gate to the fuel grate. the leftover ash was collected and weighed. The residual carbon
The biomass material is fed via the fuel inlet gate opening into content in the leftover ash is then determined using loss-on-
the reactor (Item 7 in Fig. 1). The fuel, which settles on the fuel bed ignition (LOI) approach [24]. The ash blade wiper was not operated
with grated surface (4), is heated by convective heat transfer from during the experimental runs in order to avoid disrupting the reac-
preheated air supply of an AC powered air igniter (12). The heat tions occurring in the combustion zone.
transfer increases the temperature of the fuel and initiates fuel
ignition (standard ignition); aiding the stages of drying, devolatil- 2.4. Performance evaluation
isation, and subsequent conversion of the fuel to product gas and
ash. After fuel ignition, which is evident by a sudden increase in The operational boundaries and the optimum conditions for
the combustor temperature, the igniter and the preheated air sup- converting the faecal biomass were established by varying the:
ply are shut off and air is supplied at ambient conditions via the (i) air flow rate, (ii) ignition mode, (iii) fuel pellet size, (iv) bed
primary air inlet (8). The air from the primary air inlet enters the height, and (v) fuel moisture content. Performance of the process
reactor just about the oxidation zone and moves downward was addressed by determining the MCE and gCCE.
towards the fuel bed. A rotameter (6) is fitted at the primary air The mass flow rate of the primary air (in g/s) was calculated by
inlet (3) to measure the volumetric air flow rate into the reactor. means of Eq. (1). The flue gas molar flow rate, xfg was obtained
This air stream maintains the release of volatiles from the fuel from a mass balance of nitrogen and using the concentration of
and the conversion of the formed char into non-condensable gas the constituents in the flue gas. This is expressed in Eq. (2) as g mol
and ash. Depending on the air flow, incomplete combustion may gas per second.
occur and the product gas may also include a combination of
hydrogen (H2), carbon monoxide (CO), and methane (CH4). Nitro- Air flow rate; AFRðg=sÞ ¼ X i  q=60 ð1Þ
gen (N2) from the air is also present in the outlet gas stream. The
flue gas leaves the reactor through the exhaust port positioned at xfg ðmol=sÞ ¼ AFR  VN2 air  100=MN2
the top-side of the unit (9), where it is directed to an extraction  ð100  ½CO2 ; % þ ½CO;% þ ½CH4 ; % þ ½H2 ; %Þ ð2Þ
hood via a suction fan (1). The suction fan is only operated during
fuel ignition as it facilitates the flow of preheated air into the com- where X i is the air volumetric flow rate in L/min as measured by the
bustion zone by drawing ambient air over the heated igniter sur- rotameter, q is the density of air, 1.225 g/L at standard temperature
face. The ash that is deposited during combustion is emptied into and pressure, VN2 air is the mass percentage of nitrogen in air, 76.9%.
an ash collector (14). This process is assisted by a small fan-type MN2 is the molecular weight of nitrogen, 28 g/mol.
motor (15) and a metal scraper that rubs over the grate (3). Tem- The fuel burn rate (g/s) as denoted in Eq. (3) refers to the time
perature measurements are achieved with K-type thermocouples required for complete combustion of the fuel in mixture with air
positioned at different zones (2, 9–12). and denotes the rate at which energy is released into the system.
The higher the fuel burn rate, the more efficient the system and
2.3. Experimental procedures the higher the combustion temperature, as more energy is
released.
For each experiment, 50 g of dry biomass sample (except stated Fuel burnrate; FBR ðg=sÞ ¼ Mf =tðsÞ ð3Þ
otherwise) was introduced into the reactor. This initial amount of
fuel was estimated from the maximum adult faeces generation rate where Mf is the mass of biomass consumed per experiment in g and
of 170 g/cap/day and considering a 25% total solids in human t is the duration of the experiment in s.

Table 2
Chemical and physical properties of the different biomass feedstock.

Samples Bulk density Particle Proximate analysis (wt.% dry basis) Ultimate analysis (wt.% dry basis) HHV (MJ/kg)
kg/m3 sizeb (mm)
Volatile matter Fixed carbon Ash content C H N Oa
WB 600 ± 18 8  10 98.70 ± 0.00 0.22 ± 0.00 1.04 ± 0.00 48.96 ± 0.80 6.88 ± 0.11 0.08 ± 0.00 43.04 ± 0.90 21.54 ± 0.38
SF 626 ± 5 5  10 86.77 ± 0.00 0.08 ± 0.00 13.15 ± 0.00 44.85 ± 0.10 7.24 ± 0.03 2.52 ± 0.01 32.24 ± 0.10 22.36 ± 0.06
HF 277 ± 45 10  10 85.39 ± 0.00 0.05 ± 0.00 14.56 ± 0.00 49.41 ± 0.11 7.62 ± 0.04 5.31 ± 0.02 23.10 ± 0.16 24.73 ± 0.10

HF – Human Faeces, WB – Wood Biomass, SF – Simulant Faeces; Oxygen.

a
100 – (wt.% of C, H, N and ash).
b
Average size reported.
 T. Onabanjo et al. / Fuel 184 (2016) 780–791 783

Fig. 1. Schematic diagram of the bench-scale fixed bed downdraft combustor test rig. [1] Suction Fan, [2] Exhaust Port, [3] Ash Agitator, [4] Fuel Bed (Grated Surface), [5] Air
Supply Line, [6] Rotameter, [7] Fuel Inlet Gate, [8] Primary Air Inlet, [9] Upper Combustor Temperature, [10] Lower Combustor Temperature, [11] Combustion (Bed)
Temperature, [12] Heater Temperature/Air Igniter, [13] Ash Collector and [14] Ash Rotor.

The carbon mass flow rate (mC Þ, carbon conversion efficiency The results in Table 2 show that the volatile matter and ash con-
(CCE) and modified combustion efficiency (MCE) were calculated tent of the human faeces are 85.39 wt.% db and 14.56 wt.% db
by using Eqs. (4)–(6) respectively. respectively. In the case of wood biomass, the volatile matter and
ash contents are 13.5 points higher and 13 points lower
mCðg=sÞ ¼ xfg  MC  ð½CO2  þ ½CO þ ½CH4 Þ ð4Þ
respectively. In terms of elemental composition, the contents in C
gCCE ; % ¼ ð½MC =C%fuel  C%ash Þ ð5Þ and H of the dry human faeces and wood biomass are similar.
However, differences in the nitrogen and oxygen content are
MCE;% ¼ ½CO2 =ð½CO2  þ ½COÞ ð6Þ observed; particularly in the case of the percentage composition
where C%fuel and C%ash are the percentage of carbon in the fuel and of oxygen, which is 23.10 ± 0.16 wt.% db in the case of HF and
ash respectively, and [MC] is the molecular weight of carbon, 43.04 ± 0.90 wt.% db in the case of WB. This is due to the slight dif-
12 g/mol. ferences in N and ash content of the fuels. The HHV values of both
The excess air ratio, k was calculated based on the estimated fuels are in the same order, with the human faeces having a slightly
stoichiometric air-to-fuel ratio molar fractions of the gases in the higher HHV of 24.73 ± 0.10 MJ/kg (around 13% higher than that of
flue gas and fuel burn rate as expressed in Eq. (7). WB). Comparison between the human and simulant faeces shows
that both samples present similar proximate compositions (with
k ¼ AFRreal =AFRstoic ¼ ðX i  q=Mf  AFRstoic Þ ð7Þ a percentage difference <2%). However, in the case of ultimate
analysis, slight differences in the C and N content are observed,
3. Results and discussion with a percentage difference of about 10% for C and >100% for
N. Similar to wood biomass, the HHV of the simulant faeces is
3.1. Biomass characterization slightly lower than the human faeces, an acceptable percentage
difference of 11%.
The proximate and ultimate composition of the HF, WB and SF The main differences in fuel composition and their influence on
are summarized in Table 2. Particle size, bulk density, and calorific combustion performance are discussed in Section 3.3. The
value of the fuels are also shown. composition results of the three samples are in agreement with
784 T. Onabanjo et al. / Fuel 184 (2016) 780–791

composition data of similar samples published elsewhere [23]. ignition of the simulant faeces, the results in Fig. 2a show that
Onabanjo et al. [23] showed that average composition from a set combustion temperature varied between 228 °C and 316 °C across
of 12 different human faeces sample was 51 ± 2 wt.% C, 7 ± 0 wt.% the different ERs. This range increased to 442 °C and 603 °C for the
H, 4 ± 1 wt.% N, 21 ± 3 wt.% O and 17 ± 1 wt.% ash (dry basis). In simulant faeces with booster ignition, a percentage difference of
addition, simulant faeces prepared following the same recipe used >40%. The fuel burn rate in Fig. 2b also increased from a range of
in this work had similar ultimate composition, i.e. 46 wt.% C, 8 wt.% 1.4–1.8 g/min to 2.4–3.2 g/min. In the case of the wood biomass,
H, 3 wt.% N, 30 wt.% O and 14 wt.% ash while the wood biomass, the combustion temperature was similar for both ignition types
drawn from the same pool of sample had 49 wt.% C, 7 wt.% H, within the ER range of 1–1.5. The combustion temperature had a
0.2 wt.% N, 43 wt.% O and 0.7 wt.% ash. The close compositional range of 465–592 °C for standard ignition, while 470–605 °C for
characteristics of the simulant faeces to human faeces make it a booster ignition. The fuel burn rate had a range of 1.7–2.5 g/min
suitable model fuel and the wood biomass serves as an appropriate for the case of the standard ignition of the wood biomass, and var-
reference biomass fuel. ied between 2.1 and 2.2 g/min for the booster ignition case.
Fig. 2c shows that the MCE was above 70% for all the fuel types,
3.2. Parametrisation irrespective of the ignition type, which is a good indication that the
test rig was operating within a combustion range. The combustion
Series of experiments were conducted using simulant faeces processes with booster ignition had higher MCE than their stan-
and wood biomass to understand the combustion operating range dard ignition counterparts. The use of booster ignition increased
of the test rig and to identify the optimum conditions for achieving the MCE slightly from a range of 78–85% to 83–91%, a 9% increase
complete combustion, under limited energy requirement for ignit- at the least for the wood biomass, while it was 72–80% to 78–83%
ing the fuel. The results are presented in Sections 3.2.1–3.2.4. in the case of the simulant faeces, a 6% increase at least in MCE. The
results in Fig. 2d also indicate a clear trend of the effect of ER on
3.2.1. Influence of equivalence ratio and fuel ignition carbon conversion efficiency. For all the biomass materials and
Fig. 2(a–d) presents the effect of equivalence ratio on fuel burn ignition types, there was more carbon conversion in the flue gas
rate, combustion temperature, MCE and gCCE, considering two at higher ER. The gCCE increased from the range of 14–45% to
modes of ignition (standard and booster) for the use of simulant 32–45% for the simulant faeces, indicating the significant influence
faeces and wood biomass. of the booster ignition. In the case of the wood pellet, the gCCE had
Fig. 2a and b shows the significant influence of fuel ignition on a range of 46–65% and 46–87% for booster and standard ignition
the combustion processes of simulant faeces. Under standard respectively.

Fig. 2. (a–d): Influence of equivalence ratio on: (a) combustion temperature, (b) fuel burn rate, (c) MCE, and (d) gCCE, as a function of fuel ignition mode (booster, B and
standard, S) and varying fuel types (wood biomass, WB and simulant faeces, SF).
 T. Onabanjo et al. / Fuel 184 (2016) 780–791 785

As suggested by Palmer [10] and Carvalho et al. [11], smoulder- As can be seen from Table 3, the combustion temperature
ing begins when the fuel reaches its ignition temperature; hence, increased when the bed height increased, ranging from 497 to
the insignificant difference between booster and standard ignition 655 °C. Similar trend was observed in the case of fuel burn rate,
for the wood biomass could be as a result of a fully satisfied fuel which increased from 2.15 to 3.77 g/min when the bed height
ignition process, such that the system becomes independent of increased. The increment in combustion temperature and fuel burn
the assisted heat supply. Wang et al. [25] showed that increased rate with the bed height can be attributed to the larger amount of
inflow of air enhanced flame propagation and this was attributed fuel in the reactor. Since more fuel is available for the combustion
to improved mass transfer of oxygen to the surface of the com- reaction, more heat of combustion is released into the system;
bustible material at elevated temperatures. The diffusion of oxygen hence, supporting the conversion of more fuel. Interestingly, the
is said to increase the reaction rate and consequently the genera- MCE and gCCE showed optimum values of 86% and 72% respec-
tion of heat. The improved interactions between the fuel and air tively, corresponding to a bed height of 30 mm. At bed height of
therefore enables a higher heat of combustion. Nevertheless, a high 40 mm, the reaction was observed to be operating near gasification
air flow rate also increased the loss of heat because of convective (results not shown), due to the high amount of CO (up to 10%) in the
heat transfer, and so a compromise is required. In some of the gas relative to CO2. Hence, it is possible that part of the energy
failed tests (results not shown), there was visible release of vola- released from the exothermic combustion process is used to drive
tiles that turns into flames at elevated temperatures; however, the endothermic processes of gasification of the solid residue. The
the reaction rapidly extinguishes when the air flows (>16 L/min) bed height of 20 mm on the other hand exhibited excess air and
are not favourable. In some of the instances, it was the result of reduced combustion bed temperature. Furthermore, it was
excess air flows (>16 L/min) that affect flame stability, reduces observed that at above bed height of 30 mm, there was prolonged
the combustion temperature and extinguishes any visible flame. ignition period. This is because at bed height of 30 mm, the pri-
In other cases, it was as a result of insufficient air flows to propa- mary air enters into the combustion zone just above and directed
gate the flame and heat produced. In the instances of prolonged at the solid fuel with uniform distribution of air, while at bed
ignition period, typical of standard ignition, flame propagation height of above 30 mm, the solid fuel is above the entry of the pri-
was not achieved because convective heat transfer became domi- mary air and there is poor interaction between the air and solid
nant and the release of volatiles was not accompanied by high tem- fuel. The results obtained for combustion of wood biomass at vary-
peratures. Thus, controlled air supply is important for self- ing bed height can have limited application in faeces combustion
sustained ignition, propagation and combustion of the biomass because of the significant amount of ash in the samples, as the
samples; otherwise flame extinction occurs. ash can interfere with combustion processes, especially for contin-
In this study, the optimum ER for the conversion of the simulant uous feeding processes. To this effect, a new combustor rig is under
faeces with booster ignition was 0.94–0.97, which agrees with development to manage ash accumulation.
stoichiometric combustion. This ER corresponds to air supply rate
of 14–18 L/min. The highest combustion temperature for the wood 3.2.3. Influence of fuel pellet size
biomass was achieved within the ER of 1.3–1.4, also corresponding Combustion of simulant faeces pellets of varying sizes was car-
to air supply rate of 14–18 L/min. This form of low combustion ried out to address the influence of this parameter on the perfor-
temperature, slow progressing ignition and oxidation of the fuel, mance of the rig and to establish adequate particle size for the
is classified as smouldering. While the process is occurring at faecal biomass. The simulant faeces sample was chosen for this
higher temperatures in the wood biomass irrespective of the fuel analysis due to similar physical properties and chemical character-
ignition mode and in the booster ignition of the simulant faeces, ization to human faeces, as well as the homogeneous nature of the
it is characterised by a relative long reaction time of the fuel and fuel, such that a definite pellet size can be prepared and test
having relatively low temperature in the simulant faeces (standard repeatability can be ensured. The experiments were conducted
ignition). both at standard and booster ignition modes. The influence of fuel
pellet size on combustion temperature, fuel burn rate, MCE and
3.2.2. Influence of bed height
gCCE are presented in Table 4. All these tests were conducted at
constant air supply rate of 12 L/min, as this was the optimum con-
The influence of the bed height on the performance of the rig
dition for the standard ignition of the SF sample.
was tested with wood pellets of average size of 8 mm  10 mm.
The results show that SS-2 (10 mm  10 mm square shape pel-
Combustion was carried out at fixed air volumetric flow rate of
lets) with booster ignition had the highest performance among the
16 L/min at four bed heights; namely, 20, 30, 35 and 40 mm, which
fuel pellet sizes. Thus, a combustion temperature of 520 °C and a
corresponds to 25, 50, 75 and 100 g of sample respectively (see
fuel burn rate of 2.91 g/min were achieved, which resulted in a
Table 3). Wood biomass was chosen for the evaluation of bed
MCE of 80% and gCCE of 41%. The SS-4 (5 mm  10 mm cylindrical
height influence because of the limited effect of the mode of fuel
pellets) with booster ignition had a similar performance, although
ignition and the ease of ignition. The appropriate bed height for
with a lower combustion temperature of 484 °C, higher fuel burn
the combustion of human faeces was then extrapolated from these
rate of 3.31 g/min, and similar MCE of 80% and higher gCCE of
results.
48%. The rest of the fuel sizes (SS-1 and SS-3) with booster ignition
had combustion temperature that was less than 300 °C. These sam-
ples also had extended reaction time of >20 min with values as low
Table 3 as 1.15 g/min for fuel burn rate, 72% for MCE and 18% for gCCE.
Influence of bed height for wood biomass combustion. The simulant faeces (SS-2 and SS-4) with standard ignition had
Bed Peak Fuel burn MCE gCCE Duration of lower combustion temperature than their booster counterparts,
height combustion rate (g/min) (%) (%) experiment with temperature difference of 179 °C for SS-2 and 168 °C for
(mm) temp (°C) (min) SS-4 pellet size cases. Consequently, the fuel burn rate, MCE and
40 655 3.77 71 66 26.5 gCCE had worse performance, with respective percentage
35 631 2.64 82 68 28.4 difference of 11%, 1% and 2% for the SS-2 pellet size and 29%, 13%
30 607 2.44 86 72 20.5 and 65% for the SS-4 pellet size. Comparison of the performance
20 497 2.15 80 56 11.6
of SS-1 (0.5 mm  0.5 mm) and SS-3 (20 mm  20 mm) square-
Normal ignition; Air Flow Rate – 16 L/min; Wood Biomass. shaped pellets under both ignition modes points out that the
786 T. Onabanjo et al. / Fuel 184 (2016) 780–791

Table 4
Influence of fuel pellet size and ignition for simulant faeces combustion.

Sample Pellet size (mm  mm) Ignition type Peak combustion temp (°C) Fuel burn rate (g/min) MCE (%) gCCE (%) Duration of experiment (min)
SS-1 5  5, square shape B 215 1.56 72 22 32.0
SS-2 10  10 square shape B 520 2.91 80 41 17.2
SS-3 20  20 square shape B 203 1.15 80 24 43.5
SS-4 5  10 cylindrical shape B 484 3.31 80 48 15.1
SS-1 5  5, square shape S 318 2.44 67 28 20.5
SS-2 10  10 square shape S 341 2.59 79 40 19.3
SS-3 20  20 square shape S 243 1.31 72 27 38.2
SS-4 5  10 cylindrical shape S 316 2.36 70 17 21.2

B-Booster; S-Standard.

booster ignition impeded the combustion process. SS-1 sample Table 5

achieved a much lower combustion temperature with relatively Minimum temperature required for ignition and complete conversion of moist
simulant faeces samples.
long reaction time of up to 30 min when booster ignition was
applied. This could be as a result of the inability of the preheated Moisture Combustor bed temperature
air to flow through the compact sample bed and the release of content (wt.%)
600° ± 10 °C 500 ± 10 °C 400 ± 10 °C
the volatiles being hindered. Similarly, the SS-3 fuel sample pre-
70% x x x
sented a worse performance with extended ignition period when p p
65% F = 14 L/min F = 16 L/mina x
p air p air
booster ignition was applied. The results show that the reaction 60% Fair = 16 L/min F = 16 L/mina x
p air
time of the fuel was up to 43 min, as compared to the 38 min in 55% – Fair = 16 L/min x
the case of the standard ignition. A lower combustion temperature 50% – – x
45% – – x
was also reached with temperature difference of about 30 °C when 40% – –
p
Fair = 10 L/mina
booster ignition was applied. The fuel burn rate reduced by nearly p
14% while MCE increased by 10%. This could be attributed to the Success at air flow of 16 L/min; (a) air flow rate of 14 L/min; (b) air flow rate of
16 L/min; (c) air flow rate of 10 L/min.
relative large size of the fuel pellets that necessitate the passing a
Adjusted to 10 L/min at low combustion temperature below 400 °C; – analysis
of the preheated air through the combustion zone without proper not conducted based on previous success rate; x – failed test.
interactions between the biomass material and hot air. This is evi-
dent with the relatively long reaction time of the fuel, low process
temperatures and poor gCCE. Since, MCE is a relative percentage of
CO2 to the mixture of CO2 and CO in the flue gas, the increased MCE is allowed to take place until there is an observed charring stage,
in the booster ignition case can be attributed to the relatively long that corresponds to a decline in combustion bed temperature. (B)
reaction time that permits more CO2 formation from CO in the pro- the fuel gate is opened and the sample is introduced into the fuel
duct gas and O2 in excess air. The 10 mm  10 mm square shaped bed. This causes a slight decline in combustion bed temperature.
pellets were the most suitable size for fuel conversion in the com- (C) the fuel drop disturbs the bed and stokes the ash and this
bustor test rig, as fuel size SS-3 pushes the reaction towards excess causes a sharp increase in combustion bed temperature. (D) the
oxygen range and SS-1 encourages oxygen depletion. newly introduced moist fuel begins to dry and (E) propagates into
flames after drying and volatilisation, if sufficient heat is still avail-
3.2.4. Influence of fuel moisture content able in the system. Flame propagation is accompanied by increas-
This set of experiments was carried out to investigate the min- ing rise in combustion temperature (red line) while a continuous
imum combustion temperature that is required for the combustion decline indicate the heat is continuously absorbed for drying, and
of human faeces with different contents of moisture. Onabanjo (F) fuel gate is opened and unburnt sample is recovered, if
et al. [23] have described the moisture content of human faeces available.
as ‘‘parasitic” when used as fuel; in other words, moisture content The results in Table 5 show that samples with as much as 65 wt.
can cause a significant delay or failure to ignite the fuel and other % of moisture in the fuel can be processed provided the combustion
subsequent conversion processes. This is because the energy that is bed temperature is at least 500 °C or higher at optimum air-to-fuel
released into the system is utilised in part or whole to evaporate ratio. This minimum bed temperature is critical if a self-sustained
the moisture in the sample. The moisture content can also combustion process should be achieved. In this study, 50 g of moist
compromise self-sustainability of the process. As shown in Table 5, biomass was introduced into the reactor as a batch and this means
simulant faeces with moisture content between 40 and 70 wt.% that the heat recovered from the combustion of wood biomass is
(wet basis) and minimum bed temperature between 400 and only used for drying the biomass due to the high amount of mois-
600 °C were investigated. For each test, 50–75 g of wood biomass ture in the sample. Thus, there would be need for fuel re-ignition if
that can ensure the required minimum bed temperature was new set of samples were to be introduced. In ideal systems, this
initially combusted in the reactor. At defined combustion bed tem- process will require an automated and controlled feed-in system
perature that corresponds to the point where charring starts, 50 g and rig design that would allow sufficient time for the fuel to dry
of the moist simulant faeces is then introduced into the reactor. At and progressive addition of the moist fuel, without overwhelming
the end of each experiment, a test is classified as a success, if the the on-going combustion process. Table 5 also shows that a differ-
moist faeces were completely burnt out to gas products and ash, ent air flow rate (10 L/min) was applied for the moist samples
and as a fail test, if the sample was partly converted or not all. (>55 wt.%) at minimum combustion temperature of 500 °C and
The process of fuel ignition and conversion of the simulant for the moist sample (>40 wt.%) at minimum combustion temper-
faeces with 65 wt.% moisture at 600 ± 10 °C and 500 ± 10 °C are ature of 400 °C. This is because the air flow rate of 16 L/min was
shown in Fig. 3a and b respectively. observed to rapidly reduce the combustion temperature while
The process has been classified into the following stages: (A) the lower air flow rate of 10 L/min allowed the moist fuel to com-
wood biomass is ignited using the air igniter. The ignition airline pletely dry and gradually combust, even at very low temperatures.
is switched off once ignition was achieved (blue line). Combustion These observations indicate the critical importance of controlled
 T. Onabanjo et al. / Fuel 184 (2016) 780–791 787

Fig. 3. (a and b): Combustion of simulant faeces (60 wt.% moisture): (a) at combustion temperature of 600 °C, air volumetric flow rate of 14 L/min, (b) at combustion
temperature of 500 °C, air volumetric flow rate of 14 L/min, adjusted to 10 L/min after 1500 s.
788 T. Onabanjo et al. / Fuel 184 (2016) 780–791

air supply and minimum temperature of the bed. In the experi- dry sample that corresponds to a bed height of 30 mm was intro-
ments with success, the moist samples were completely converted duced into the reactor. The result findings are indicated in Table 7.
to ash with <1% of residual carbon. The results in Table 7 show that the combustion temperature
for the dry human faeces increased from 431 to 558 °C while the
fuel burn rate increased from 1.53 to 2.30 g/min, with increasing
3.3. Human faeces and combustion performance air supply rate, corresponding to ER of 0.86–1.12. These results
show that the combustion processes for the human faeces were
Based on the Bristol Stool Chart (BSC), human faeces can be cat- close to stoichiometric combustion, an ideal condition where the
egorised by physical appearance into 7 types of stool. The stool carbon and hydrogen in the faeces mixes with sufficient air and
types vary from ‘Type 1’ stool, classified as ‘separate hard lumps, reaches a maximum temperature and complete conversion to
like nuts’ to ‘Type 7’ stool, listed as ‘watery, no solid pieces’, as H2O and CO2. The optimum combustion temperature and fuel burn
shown in Table 6 [26]. The 2–4 type stools were commonly found rate were obtained at the highest air supply rate of 18 L/min, due to
among the samples collected. increased oxidation of the fuel. While the highest burn rate was
The various stool types also exhibited different moisture con- achieved at air flow rate of 18 L/min, the highest MCE and gCCE
tent levels, with the Type 1 having the least moisture content were obtained at air flow rate of 10 L/min, a range of 77–89% for
(50%) and Type 7 having the most (>80%). However, some of MCE and 53–82% for gCCE. The improved combustion efficiencies
the samples in the study had definite forms as described in the at low air supply rates can be attributed to the long residence time
BSC (Fig. 4a) while other samples were a mixture, including Type of the fuel (>30 min) that provides more time for the sample to be
2 and Types 4–6 stool characteristics, in some instances (Fig. 4b). completely burnt out to CO2 and H2O.
Because of the heterogeneous dispersion of moisture content Among the different operating conditions, fuel ignition
across each sample and the mixture of stool types even within a sequence and fuel characteristics had a significant influence on
given sample, the various classes of stools are observed to exhibit the combustion processes of the dry human faeces. Fuel ignition
different drying profiles (Fig. 4c and d) and the overall moisture is a critical parameter that affects the amount of heat released into
content level in the sample did not defer significantly. For instance, the system, as it has a direct influence on combustion temperature
the sample in Fig. 4a and b had 65 and 76 wt.% moisture, and fuel burn rate. Monhol and Martins, [20] exposed faeces to
although the Type 6 section of the mixed sample 86 wt.% heat influx of 30 kW/m2 from a radiant cone heater at elevated
moisture. temperatures of 570 °C and achieved a combustion temperature
The type 1–2 stools were observed to dry completely as a lump of about 885 °C. Their studies showed that the ignition tempera-
with minimal ‘shrinkage’, whereas the type 5–7 stools dried typi- ture of human faeces is about 220–375 °C and exhibits a heteroge-
cally as sheets or flakes of solids with high ‘shrinkage’, that is a neous behaviour. In the case of an ‘heterogeneous’ fuel ignition,
large reduction in volume of the sample in relation to the original there is direct interactions of oxygen and the organic matter on
volume. This could because the type 1–2 stools have a visible com- the surface of the fuel; however, ‘homogenous’ ignition occurs in
pact structure, hold less visible air spaces, and relatively low fuel the gas surrounding the fuel [11]. In this study, the combustion
moisture. The type 6–7 stools are not so compact and can hold high of faeces can be described to exhibit smouldering ignition with
volume of gases that disperses during the dry stages. Hence, complex heterogeneous reactions that transit into flame propaga-
sample homogeneity was an important aspect of the faeces com- tion. This observation holds for faeces with ‘booster’ ignition,
bustion tests. because the heat flux from the air igniter increases the tempera-
To minimise fuel variations, all the collected samples were ture of the fuel directly and enhances the thermal decomposition,
mixed together to obtain a relatively uniform consistency and such that drying, pyrolysis and release of gaseous volatiles are fas-
distribution of undigested foods such as vegetables and grains. tened for oxidation to occur. In the case of the standard ignition,
Nevertheless, a definite pellet size could not be achieved due to the fuel is gradually heated and the volatiles are released, causing
mass shrinkage of the sample during drying and non-uniform in smouldering ignition to be dominant without flame propagation,
drying characteristics. As such, the samples were bulk dried at resulting in a low combustion temperature, as described in Sec-
minimal oven temperature of 45 °C and crushed to an estimated tion 3.2.1. Similar to the simulant faeces, the maximum combus-
mean pellet size of 10 mm. Based on the experimental outcomes tion temperature achieved under standard ignition of the dry
in Sections 3.2.1–3.2.4, the combustion analyses of dry human fae- faeces was at most 300 °C at optimum air flow rate of 16 L/min
ces were conducted at air flow rate of 10–18 L/min using booster (results not shown). Since, minimal energy requirement is one of
ignition. For consistency and appropriate comparison, 50 g of the the design considerations for a self-sustained energy conversion
system for the NMT, the input power is estimated at 1.2 kW for
booster ignition and 2.9 kW for standard ignition. During the
Table 6
Bristol stool chart [26] with broad classification of moisture content from study. booster ignition of the dry faeces, the air igniter was operated at
heater temperature between 620 °C and 627 °C for 5–6 min. This
Type 1 ‘‘separate hard lumps, like nuts” 50%
includes the time required for the heater to reach >600 °C and
the period in which the suction fan is used to draw ambient air
Type 2 ‘‘sausage-shaped but lumpy” 50–65%
across the heated igniter surface until fuel ignition is achieved in
Type 3 ‘‘like a sausage but with cracks
the combustion zone.
on the surface” Fuel characteristics is another critical parameter that influences
Type 4 ‘‘like a sausage or snake smooth 65–80% the performance of the combustion process. In Section 3.1, it was
and soft” mentioned that there are slight differences in the proximate com-
Type 5 ‘‘soft blob with clear cut edges” positions of WB and HF. The volatile matter content of the dry WB
was up to 13% higher than those of HF with insignificant amount of
Type 6 ‘‘fluffy pieces with ragged edges, >80% ash. These slight differences had influence on combustion temper-
a mushy stool”
ature (<600 °C) as higher values were reached with WB. Volatile
Type 7 ‘‘watery, no solid pieces” matter is readily given off during combustion after drying of the
sample, and can contribute to spontaneous combustion of the fuel,
as it encourages fuel oxidation [27]. Ash on the other hand is
 T. Onabanjo et al. / Fuel 184 (2016) 780–791 789

Fig. 4. (a–d): Human Faeces: (a) Type 2 fresh sample (as-received basis, prior to drying. (b) Mixture of Type 2, 5 and 6 fresh sample (as-received basis, prior to drying). (c)
Type 2 sample (after drying). (d) Mixture of Type 2, 5 and 6 sample (after drying).

Table 7
Combustion performance at varying air supply rate.

Sample Air volumetric flow rate (L/min) Ignition type Peak comb. temp (°C) Duration of experiment (min) Fuel burn rate (g/min) MCE (%) gCCE (%)
HF 18 B 558 21.8 2.30 77 55
HF 16 B 542 25.1 1.99 83 63
HF 14 B 547 24.7 2.02 87 68
HF 12 B 431 26.1 1.91 87 57
HF 10 B 435 32.7 1.53 89 82

largely inert and do not contribute to combustion temperature, but process. In these instances, the inlet of the air igniter and the pri-
rather absorbs heat during combustion. Kaymakci and Didari [27] mary air flow were found to have a blockage. Since the test rig
showed that ash can reduce the spontaneous ignition of coal and has a downdraft configuration, the build-up of ash also disrupts
attributed this to certain components. Hence, a high composition the passage of air and product gas, thereby limiting the reaction
of ash in the samples can reduce the temperature that could have considerably. Therefore, the design and development of a fuel con-
been given off by the process. In smouldering combustion, ash can version system for application in sanitary systems require an effi-
act as a porous medium to aid the flow of oxidant and increases the cient ash control system.
surface area for the fuel and air to interact. It is also suggested that Another aspect of fuel composition is the elemental con-
ash act as thermal insulator to conserve the heat generated and to stituents of C, H, O, N. Rose et al. [1] showed that human faeces
prevent heat loss into the environment [8]. Both phenomena were contain large amount of undigested cellulose, vegetable fibres
observed in the conversion of HF and SF. The accumulation of ash and pentosane that varies with dietary intake and depends on
on the combustion bed appears to serve as a heat sink thereby the fraction of absorbed soluble carbohydrate. Here, readily avail-
allowing combustible material to have sufficient time to be heated able carbohydrate such as starch and sugars increases faecal
and to follow through with the processes of combustion. However, energy and conversely for unavailable carbohydrates. A high fibre
excessive accumulation of ash appears to disrupt the combustion content in the faeces from undigested plant matter also increased
790 T. Onabanjo et al. / Fuel 184 (2016) 780–791

the calorific value of the fuel. In this study, the fixed carbon of the characteristics, fuel composition, mainly moisture content of the
sample was in insignificant amount, suggesting readily available incoming samples, and the efficiency of the heat recovery system,
carbon in the fuel that corresponds to higher HHV of the fuel. Since, further investigation will aim to establish the appropriate
HHV is the gross amount of heat produced on complete combus- trade-off limits for drying and energy recovery under the
tion of a given mass of fuel, the relatively high HHV of the faeces consideration of different stool types, moisture content and drying
can be related to the elemental composition of the fuels, mainly characteristics.
hydrogen:carbon weight (H/C) and oxygen:carbon weight (O/C)
ratios. The HF have a slightly higher H/C, while lower O/C than
the WB, respective values of 0.15 and 0.47 for HF and 0.14 and 4. Conclusions
0.88 for WB. The slightly higher hydrogen per carbon contributes
to a lower oxidation state of the fuel and more energy release. Experimental investigation of faeces combustion was carried
The reduced oxygen per carbon also promote more oxidation of out using a bench-scale downdraft combustor test rig. Air flow
the fuel, hence more CO2 formation, as shown by the MCE (<80%) rate, fuel pellet size, bed height, and fuel ignition mode were varied
for the HF samples. Although, the dry HF sample had a relatively in order to establish the combustion operating range of the test rig
high HHV than WB, the dry HF was less an ignitable material than and the optimum condition required for using converting the fae-
wood biomass and this can be related to the different volatile and cal biomass. Wood biomass and simulant faeces samples were also
ash contents. Finally, the HF sample had the highest N content; but used to preliminary evaluate the performance of the equipment.
this is irrelevant for this study, as N does not take part in combus- Performance evaluation was carried out on the basis of combustion
tion except at very high temperatures (>1800 °C), and this is out- temperature, fuel burn rate, MCE and carbon conversion efficiency.
side the range of this study. The experimental results show that dry human faeces have a
Beyond the analyses on the dry human faeces, attempt was higher energy content than wood biomass on a dry basis. Fuel igni-
made to combust moist (as received basis) human faeces at opti- tion had a significant influence on the maximum combustion tem-
mum air flow rate of 16 L/min and minimum bed temperature of perature in the combustor test rig. Here, the exposure of elevated
600 °C, using Type 2 (62 wt.% moisture) and Type 6 (88 wt.% mois- temperatures of up to 600 °C for a very short duration of about
ture) stool types. All the six attempts had no success rate, largely 60 s (booster ignition) was sufficient to ignite the fuel and push
due to the non-definite form of the moist faeces. Although, the the reaction towards flame combustion. Fuel burn rate of 1.5–
Type 2 faeces was able to hold a form, it was held bound by 2.3 g/min was achieved at air flow rate of 10–18 L/min. Increased
mucous as the sample was introduced into the reactor. The heat air flow rate improved combustion bed temperature, however
generated from the wood biomass combustion was however only reduced the reaction time and consequently the carbon conversion
sufficient to partly dry the moist faeces, without fuel ignition. and combustion efficiencies of the fuel. Further work will be
Further work will therefore aim at creating a homogenous faeces required at creating a homogenous faeces pellet size and to estab-
pellet size. lish the appropriate trade-off limits for drying and energy recovery
The work presented in this study is expected to influence the under the consideration of different stool types, moisture content
design, development and optimisation of an appropriate energy and drying characteristics. This can improve the drying processes
conversion system for the NMT. The toilet is intended to safely of faeces combustion significantly and can ensure the successful
treat human waste onsite under limited energy requirement and development of a self-sustained energy conversion and recovery
achieve self-sustained operation with no external power supply. systems for the NMT and similar sanitary solutions.
Hence, the temperature between 431 °C and 558 °C obtained at
air flow rate of 10–18 L/min and the minimum combustion tem- References
perature of about 500 °C for combusting dewatered, partly dried
faeces suggests the need for a well-insulated system that can [1] Rose C, Parker A, Jefferson B, Cartmell E. The characterisation of faeces and
urine; a review of the literature to inform advanced treatment technology. Crit
ensure minimal heat loss into the environment. Also, of high Rev Environ Sci Technol 2015;45(17):1827–79.
importance is the need to ensure the safety of the users, thus par- [2] WHO, UNICEF. Progress on sanitation and drinking water – 2015 update and
ticular attention is being given to the choice of material for the MDG assessment. Geneva, Switzerland: WHO Press; 2015.
[3] Danso-Boateng E, Holdich RG, Shama G, Wheatley AD, Sohail M, Martin SJ.
design of the new combustor to ensure it retains heat, unlike cur-
Kinetics of faecal biomass hydrothermal carbonisation for hydrochar
rent combustor rig that dissipates heat to the surroundings. The production. Appl Energy 2013;111:351–7.
reported fuel burn rates of 1.53–2.30 g/min at air flow rates of [4] United Nations. Department of Economic and Social Affairs, Population
10–18 L/min imply that dry human faeces from about 55–82 per- Division. World Population Prospects: The 2015 Revision, Key Findings and
Advance Tables. Working Paper No. ESA/P/WP.241; 2015.
sons can be treated every day, assuming an average faeces genera- [5] Global Development Program (GDP). Water, sanitation and hygiene: strategy
tion rate of 40 g/cap/day (dry basis). Rose et al. [1] indicated a wide overviewAvailable from: <http://www.gatesfoundation.org/What-We-Do/
range of faeces generation rate of 51–796 g/cap/day (as received Global-Development/Water-Sanitation-and-Hygiene>2013. Retrieved May
25, 2016.
basis), with a median value of 128 g/cap/day. In low- and high- [6] Parker A. Membrane technology plays key role in waterless hygienic toilet.
income countries, respective median fresh faeces generation rates Membr Technol 2014;2014(12):8.
of 250 and 126 g/cap/day were reported. Total solids were in the [7] Nussbaumer T. Combustion and co-combustion of biomass: fundamentals,
technologies, and primary measures for emission reduction. Energy Fuels
range of 11–34% with a median value of 25%. Thus, at the upper 2003;17(6):1510–21.
limit of faeces generation rate, the fuel burn rate still corresponds [8] Rein G. Smouldering combustion phenomena in science and technology. Int
to the daily dry faecal treatment of 11–16 persons, and this meets Rev Chem Eng 2009;1:3–18.
[9] Malow M, Krause U. Smoldering combustion of solid bulk materials at different
one of the objectives of the NMT to treat faeces of single-household volume fractions of oxygen in the surrounding gas. Fire Safety Sci
use (equivalent to 10 people). For optimal continuous operation 2008;9:303–14.
and low fuel burn rate, equivalent to 10 people, a miniature [10] Palmer KN. Smouldering combustion in dusts and fibrous materials. Combust
Flame 1957;1(2):129–54.
combustion system is being developed. An automated ash control
[11] Carvalho ER, Veras CAG, CarvalhoJr JA. Experimental investigation of
system is also being developed to ensure that ash accumulation smouldering in biomass. Biomass Bioenergy 2002;22(4):283–94.
does not hinder the faeces conversion processes. A drying [12] Pironi P, Switzer C, Rein G, Fuentes A, Gerhard JI, Torero JL. Small-scale forward
chamber with multiple capabilities as a pelletiser is planned to smouldering experiments for remediation of coal tar in inert media. Proc
Combust Inst 2009;32(2):1957–64.
ensure extended operation and continuous feed and drying of [13] Rein G, Cleaver N, Ashton C, Pironi P, Torero JL. The severity of smouldering
samples. Because energy recovery will depend on drier peat fires and damage to the forest soil. Catena 2008;74(3):304–9.
 T. Onabanjo et al. / Fuel 184 (2016) 780–791 791

[14] Rashwan TL. Self-sustaining smouldering combustion as a novel disposal [21] Monhol FAF, Martins MF. June. Ignition by thermal radiation of polyethylene
destruction method for waste water biosolids Doctoral dissertation. The and human feces combustible wastes: time and temperature to ignition.
University of Western Ontario; 2015. Advanced materials research, vol. 911. p. 373–7 [June].
[15] Pironi P, Switzer C, Gerhard JI, Rein G, Torero JL. Self-sustaining smoldering [22] Pollution Research Group (PRG). Selection of synthetic sludge simulant for the
combustion for NAPL remediation: laboratory evaluation of process sensitivity bill and Melinda gates foundation’s reinvent the toilet fair: India 2014. Report
to key parameters. Environ Sci Technol 2011;45(7):2980–6. by pollution research group, University of KwaZulu-Natal, South Africa; 2014.
[16] Yermán L, Hadden RM, Carrascal J, Fabris I, Cormier D, Torero JL, et al. [23] Onabanjo T, Patchigolla K, Wagland ST, Fidalgo B, Kolios A, McAdam E, et al.
Smouldering combustion as a treatment technology for faeces: exploring the Energy recovery from human faeces via gasification: a thermodynamic
parameter space. Fuel 2015;147:108–16. equilibrium modelling approach. Energy Convers Manage 2016;118:364–76.
[17] Wall H, Gerhard J, Fabris I, Cormier D, Cheng YL, Torero JL. Investigation of self- [24] Zhao M, Han Z, Sheng C, Wu H. Characterization of residual carbon in fly ashes
sustaining smouldering of faeces: key parameters and scaling effects. In: from power plants firing biomass. Energy Fuels 2013;27(2):898–907.
Dynamic ecolibrium: sustainable engineering society conference (SENG 2015). [25] Wang JH, Chao CYH, Kong W. Experimental study and asymptotic analysis of
Engineers Australia; 2015. p. 113. horizontally forced forward smoldering combustion. Combust Flame 2003;135
[18] Ward BJ, Yacob TW, Montoya LD. Evaluation of solid fuel char briquettes from (4):405–19.
human waste. Environ Sci Technol 2014;48(16):9852–8. [26] Lewis SJ, Heaton KW. Stool form scale as a useful guide to intestinal transit
[19] Afolabi OO, Sohail M, Thomas CPL. Microwave hydrothermal carbonization of time. Scand J Gastroenterol 1997;32:920–4.
human biowastes. Waste Biomass Valorization 2015;6(2):147–57. [27] Kaymakci E, Didari V. Relations between coal properties and spontaneous
[20] Monhol FAF, Martins MF. Cocurrent combustion of human feces and combustion parameters. Turk J Eng Environ Sci 2001;26(1):59–64.
polyethylene waste. Waste Biomass Valorization 2015;6(3):425–32.