Woodstoves : theory and applications in developing countries

Woodstoves : theory and applications in developing countries

Citation for published version (APA):

Bussmann, P. J. T. (1988). Woodstoves : theory and applications in developing countries. Technische Universiteit Eindhoven. https://doi.org/10.6100/IR291952

DOI:

10.6100/IR291952

Document status and date: Published: 01/01/1988 Document Version:

Publisher’s PDF, also known as Version of Record (includes final page, issue and volume numbers) Please check the document version of this publication:

• A submitted manuscript is the version of the article upon submission and before peer-review. There can be important differences between the submitted version and the official published version of record. People interested in the research are advised to contact the author for the final version of the publication, or visit the DOI to the publisher's website.

• The final author version and the galley proof are versions of the publication after peer review.

• The final published version features the final layout of the paper including the volume, issue and page numbers.

Link to publication

General rights

Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights. • Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain

• You may freely distribute the URL identifying the publication in the public portal.

If the publication is distributed under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license above, please follow below link for the End User Agreement:

www.tue.nl/taverne

Take down policy

If you believe that this document breaches copyright please contact us at:

openaccess@tue.nl

providing details and we will investigate your claim.

~

­

l

WOODSTOVES

theory and applications in developing co u ntries

Woodstoves

CIP-DATA KONINKLIJKE BIBLIOTHEEK. DEN HAAG

Bussmann, Paulus Josephus Theodorus

Woodstoves : theory and applications in developing countries / Paulus josephus Theodorus Bussmann. - [S.l. : s.n.]

Thesis Eindhoven. Wi th ref.

ISBN 90-9002505-7

SISO 646.1 UDC [683.94:691.11](1-7721773)(043.3) Subject headings: woodstoves / stove designing

WOODSTOVES

THEORY AND APPLICAllONS IN DEVELOPING COUNTRIES

PROEFSCHRIFT

ter verkrijging van de graad van doctor aan de

Technische Universiteit Eindhoven, op gezag

van de rector magnificus, Prof. Ir. M. Tels,

voor een commissie aangewezen door het college

van dekanen In het openbaar te verdedigen op

dinsdag 1 november 1988 te 16.00 uur

door

PAULUS JOSEPHUS THEODORUS BUSSMANN

Dit proefschrift is goedgekeurd door de promotoren: Prof.dr.ir. G. Vossers en Prof.ir.

J.

Claus Co-promotor: Dr. K. Krishna Prasad

to all who have contributed

aan allen die hebben bijgedragen

1. Scope of the work 1.1. General 1.2. Stove testing 1.3. Open fires 1.4. Shielded fires 1.5. Stove designing

2. Stove te'sting, experimental set-up and procedures 2.1. Introduetion

2.2. The W.S.G. test procedure 2.3. The power output

2.3.1. Nominaland average power 2.3.2. The time dependent power 2.3.3. The maximum power

2.3.4. The minimum power 2.3.5. The design power 2.4. The efficiency

2.5. The

co-co

₂ ratio

2.6. The specific consumption 2.7. The VITA procedure 2.8. Concluding remarks

3. Open fires: Experiments and theory 3.1. Introduetion

3.2. Experimental parameters and procedure 3.3. Experimental results and discussion

3.3.1. General

3.3.2. The power range of the fire on a grate 3.3.3. The wood species

3.3.4. The moisture content

3.3.5. The size of the wood blocks 3.3.6. The position of the pan

1 2 5 6 8 11 11 12 12 15 18 20 24 25 27 30 32 33 35 37 38 38 39 41 43 45 46

3.3.7. The use of a grate 3.4. A theory for open fires

3.4.1. General 3.4.2. The fuelbed 3.4.3. The fire plume 3.5. The open fire with a pan

3.5.1. General 3.5.2. Radlation 3.5.3. Convee ti on

3.5.4. Results and discussion 3.6. Concluding remarks

4. Parameter analysis of simple single pan-hole stoves 4.1. Introduetion

4.2. Flow resistance and stove performance 4.2.1. Shielded fire experiments 4.2.2. Model calculations

4.2.3. Pressure measurements a long the flow pa th 4.3. The combustion chamber height

4.4. Concluding remarks

5. A consumer oriented design procedure for woodburning stoves 5.1. Introduetion

5.2. Objectives

5.3. Social (customer ?) considerations

5.3.1. The needs

5.3.2. The local resources 5.3.3. Health and safety aspects 5.3.4. Comfort aspects

5.4. Development considerations 5.4.1. General

5.4.2. The wood energy market 5.4.3. The wood stove market

5.5. Selection of the conceptual designs 5.6. Process design

5.6.1. The physical quantities

49 51 51 53 55 63 63 63 64 69 71 75 76 76 82 91 95 98 101 105 106 106 109 116 117 118 118 119 120 124 125 126

5.6.2. Oimensioning 5.7. Machanical design 5.8. Production design 5.9. Manufacturing 5.10. Marketing 5.11. Discussion

6. Summary and Conclusions

Appendix

List of symbols Raferences Samenvatting

the stove elements 132

144 147 HB 151 153 157 163 165 167 173

1

CliAPTER 1 SCOPE OF TIIE WRK

1.1. General

It is an established fact that vast areas of woodland in developing countries have been deforested. For countries at the fringe of a desert. this situation has led. and will lead. to an irreversible decrease of the area available for a productive ecosystem. A variety of actions have been proposed in order to ease the pressure on the woodland. lmproved agr icultura1 ha bits. improved forest management. reafforestation. fuel switching, and improved wood stoves will all play a role. The right mix

of actions will obviously depend on local conditions and can only be

undertaken on a factual or analytica! basis.

Biomass products are the main energy souree for both dornestic and smal! scale industrial energy needs in many developing countries. Data from more than fifteen UNDP/World Bank country assessment reports show the

household sector accounting for 30% to 99% of total energy consumption

(Leach

&

Gowen 1987). In 1981, FAO estimated, that nearly 1 billion

people are living in areas with acute fuel scarcity (FAO 1981). The

direct link between wood use for dornestic energy demands and depletion

of forest resources. is most apparent near the urban eentres where over-exploitation has led to large environmental problems. The World Bank report on desertification in Africa therefore rightly states that the challenge is not "stopping the deserts advance from the North", but effectively rnanaging renewable resources south of the desert (World

Bank 1985), not least. because wood will continue to play a major role

in satisfying the urban dornestic energy demands.

Supporters and practitioners of Appropriate Technology were the first to recognise the potentlal of wood saving stoves. Appropriate Technology failed, however, to set-up any significant stove project of sizable magnitude. Barring a few notabie exceptions. the field of Appropriate Technology suffered from a lack of professionallam and did not even

2 Chapter 1

this attitude are many. In descrihing a stove project in Mali, Templeman (1982) states that: "the improved stove is simple enough that one need only to see it to be able to reproduce it. It can be copied and passed from one neighbour to the next. In this way the stove can be adopted on a massive scale and spread easily by diffusion". Needless to say that the mass dissemination never took place.

The physics of woodstoves are complex due to the combustion process involved. Underestimating the problem wil! easily lead to design errors. The Netherlands Minister for Development Cooperation therefore

established the Woodburning Stove Group (WSG) to provide scientific support to field groups on the design and dissemination of improved stoves. The present work was done within the frameworkof the activities of the WSG. It contributes to the creation of the body of reliable knowledge on stoves. Modelling (of the cooking task, of the heat transfer in open/shielded fires and also of the design process) is the keyword in the work. Modelling is what is required if one is to

contemplate the availability, over a period of twenty years, of stoves with superior performance, to a billion people, living in diverse places with varying local conditions. (Krishna Prasad

&

Verhaart, 1983).

Ghapter 2, 3 and i of the thesis have been built around four publisbed papers on the work done in the Netherlands. Chapter 2 describes the experimental set-up and procedures, chapter 3 presents the work done with the open fires (the traditional stove) and chapter i gives a parameter analysis of a simple improved stove. The interaction between the laboratory results and data from the field, is shown in chapter 5 in which a design procedure is presented and discussed. Chapter 6 finally summarizes the main conclusions.

In the rest of this chapter short introductions are given on the work presented in the respective chapters.

1.2. Stove testing

Modelling the cooking task is required in testing the performance of new stoves and comparing the results with traditional stoves. Without

Scope of the work 3

evaluate the achievements of stove programmes. However, stove testing is still not cornmon practica in many dornestic energy programmes. To fill the gap in knowledge, projects have attempted to monitor the energy consumption at the family level. This monitoring often took the form of a census, using dubious questions so worded as to encourage the desired

answers (namely that the new stoves were superior and save 50% energy or

more). It also often ignored the complex interrelationship among the

variables determining the energy consumption (Leach

&

Cowen 1987). An

example of these dependencies is given in figure 1.1 where the wood consumption per capita is shown as a function of the family size. The

data were gathered in Niamey by Strasfogel (1983). The strong dependenee

of the wood consumption per capita on family size is clear. ' 3.5 0. 25 + ---,,---Fam.Size Oo 0 0~---~---~----~ _{0 10 20} Family size (-)

Fig.1.1: Wood consumption per capita per day as a function of the famlly size

By "the performance of a stove" is meant the whole set of relationshlps among all the varlables which determine the behaviour of the stove in

practice. This definltion is much broader than, and should not be

confused with, that of the efficiency. The latter is directly related to the energy balance measured under well defined conditions and is thus only one (important) aspect of the performance.

As a matter of fact, the questlon about the performance of the open fire has been the startlng point for the WSG. The work of the group soon

4 Dlapter 1

revealed that reliable and comparable experimental data are only obtained when stoves are tested according to a fixed test scheme. The

need for standardization also became apparent from field data. An

example is given in table 1.1. where data are presented from tests on the open fire (Gill 1985). The quoted efficiencies for the open fire

range from 3% to 15% !

Cooking metbod Author

Open fire IAE '78

Open fire Gamser '79

Open flre Hammer '80

Open flre Morgean, Moss '80

Open fire Openshaw '77

Open flre+rocks VITA '80 Primitlve stoves Desch '73

Three stones Ki-Zerbo et.al. '79

Stoves, Flre pl. Knowland et.al. '79

Three stones Moss et al. '80

Three stones Club du Sahel '79

Open fire Dunkerley '79

Three stones Mnzava '80

Two stones Weir et.al. '80

Open fire Makhijani et.al. '75

Three stones Morgan et.al. '79

Open fire Coldamberg et.al. '79

Some stones Arnold '79

Trad.Open fire ROCAP '79

Pot with legs Best '79

Open fire Reveil '78

Traditional Floor '77

Open fire Argal '78

Traditional way Spears '78

Open fire Frida '80

~quoted Souree of ~ data

(%)

5 - 15 no souree no souree no souree no souree 10 6 - 8 5 - 10 7.5 no souree wasteful no souree 10 no souree 3 - 8 no souree 2 - 10 no souree 7 - 8 no souree 3 - 8 no souree 10 no souree 10 - 12 no souree

4 - 5 Water sirnmaring test

5 Comparison India-US 5 - 10 Coldamberg '79 5- 10 Franklin et.al. 8 Makhijani,Brown,Floor <10 no souree 2.5 boiling test 10 no souree 8 Developp.Voltaique'76

1/4 ~ karosene stove,no souree

1/5 ~ karosene stove,no souree

2.5 IAscough

Table 1.1 Quotedor measured heat transfer efficiency of the open fire in the field. (Source: Gill 1985)

The WSG made several contributions to the work aimed at establishing standard test procedures. The group showed the need for eliminating all variables in stove testing not directly related to construction (Claus et al. 1982). To people in the field, however, the relationship between laboratory results and actual cooking thereby became increasingly remote. This state of affairs has blocked the necessary communication between the laboratory and the field ever since.

Scope of the 11.10rk 5

From 1981 onwards attempts were made to resolve the controversy, resulting in a provisional standard in stove testing that aimed at ensuring not only technica! performance but also the socio-economie and commercial viability of stoves (VITA 1982). The VITA procedure is clearly a compromise and is far from ideal. The methodology fails to give insight into the reasons why stoves perform differently and it neglects the existing relationships among the parameters which determine the fuel savings of a stove. In designing, these relations are of the utmost importance. The experimental set-up and procedures used in the present work therefore had to deviate from those proposed by VITA. The set-up, together with a first discussion on the test procedure. was publisbed as a part of an artical in the Journal of the Indian Academy of Sciences (Bussmann et al. 1983). A more complete discussion of the test procedure was given in a paper presented at the Third E.C. Conference on Energy from Biomass (Bussmann et al. 1985) In fact. the latter material forms the main part of chapter 2. Probably the most practical advantage of the WSG procedure is that it glves highly reproduelbie results (figure 2.13)

1.3. Open flres

Most of the earlier work of the WSG (of which an important part is presented in chapter 3) has been restricted to open fires for three important reasons (Krishna Prasad 1980):

i. lt appears that the traditional stoves of the overwhelming majority

of the poor people in this world are open fires or their close relatives. Thus, in most cases, improving the traditional cooking equipment means improving upon the performance of the three stone or similar fire. Consequently the performance of the open fire plays a key role in the efforts to disseminate improved woodstoves. ii. Water boiling tests performed on open fires in the WSG laboratory.

showed surprisingly high efficiencies. Visser (1982) showed that the traditional stove, when carefully tended, performed much better than many "improved stoves of the first generation".

Verhaart (1983) elaborated on this theme in hls artiele "Making Do

6 Chapter 1

the design of improved stoves.

111 If we are able to understand the open fire it will provide very

useful guidelines for designing more efficient stoves.

The work withopen fires has been both experimental and theoretica! in nature. It was conducted both to observe and to explain the effect of varying a number of parameters which could be expected to affect the behaviour of the fire. No attempt was made to look into the precise nature of the complex chemistry of combustion. Time and financlal constraints did not allow for any such in depth study. It was therefore only aimed at getting a qualitative picture of the processas involved. The attitude taken towards the research work. was in full agreement with the communis opinio expressed by Emmons (1983), who stated that the scientific understanding of the combustion of wood is not yet mature enough to provide quantitative prediction, but is very useful for providing qualitative predictions.

The results of the work withopen fires, presented in chapter 3, is based on articles publisbed in the Proceedings of the Seventh IHTC

Conference (Bussmann

&

Krishna Prasad 1982) and the Proceedings of the

Indian Academy of Sciences (Bussmann et al. 1983). Two, more recent contributtons have been added. First, the workof Herwijn (1984), who performed experiments with the open fire within the framework of his final thesis for the Department of Applied Physics. Herwijn measured:

i. the flame height of open fires withand without a grate (data

presented insection 3.4.3), and;

ii. the radlation from the fuelbed and the flames with an elliptical radiative heat flux meter (section 3.4.2).

Second an improved analysis of the experiments on the open fire with a parametrically varled fuelbed-pan distance has been included

(section 3.3.6 and 3.5.4.).

1.4. The shielded fire

The work on open fires received a logica! follow-up in the work on shielded fires (see figure 4.1). In its basic form, the shielded fire consists of two parts; the combustion chamber with lts top plate, and

Scope of the 100rk.

the pan shield. The function of the shield is three-fold:

i. it protects the fire against wind;

ii. it directs the hot cernbustion gases towards the pan surface; and iii. it offers the possibility of centrolling the cernbustion air flow.

7

With a shielded fire, Visser (1981) obtained efficiencies of more than

50%. Moreover, he was also able to effectively control the power output.

In 1981, Visser's shielded fire completely differed from other improved stove models. The latter normally had more than one pan-hole and were made of clay (see figures 5.1 and 5.2). However, the need fora new

improved stove concept was evident. Stove projects, five to ten years old then, had tried to disseminate the multi pan-hole stoves in large numbers, bu't the resul ts obtained we re marginal. The dissemination could only be sustained by a high capita! input from donor countries. The main obstacles encountered were:

i. a limited production output;

ii. high production costs;

iii. limited possibilities for quality control;

iv. technica! constraints;

v. high maintenance requirements; and

vi. no selling facilities through local markets.

A shielded fire type stove could meet most of these problems. Using existing private metal workshops to produce such portable metal stoves, offered much better prospects because it would reduce the main function of stove projects to stove promotion and quality control, and leave the actual production to private enterprise. However the shielded fire also has a number of important drawbacks. Delepeleire

&

Christiaens (1983) suggest first, that high efficiencies only result when the width of the gap between the pan and the shield is extremely small (less than 5 mm). Such close tolerances are rarely feasible in the production facilities accessible to cookstove fabrication. Second, the device, in its original form, accepted only one pot size. This feature will dissuade many

prospective users from consiclering the device. Finally, the high efficiencies were often obtained at the expense of cernbustion quality.

In view of the above, several series of experiments were performed to provide baseline data on the possible effects of different design

8 Oiapter 1

process of design optimization for diverse local conditions in

developing countries. A simple theory, based on the balance between the available draught and the flow resistance, qualitatively explains the observed results.

The work with shielded fires is presented in chapter 4 and is based on a

paper presented at the Sth IHTC Conference in San Fransisco (Bussmann

&

Krishna Prasad 1986). Section 4.2.3, on the pressure measurements. and some data contained in section 4.3, on the influence of the cambustion chamber height on the cambustion quality, have been added afterwards.

1.5. Stove designing

Over the years. for many field workers involved in the dissemination of stoves, the value of the laboratory results remained suspect. It became clear that the R&D work could only become rapidly effective through direct field involvement. A close working relationship was therefore established between the stove branch of the "Mission Forestiere

Allemande" (MFA), the "Ins t i tut Voltaique de 1 'Energie" ( IVE) in Burkina

Faso (Upper Volta) and the WSG. In the period from December 1982 to March 1983, a research programme was carried out in Ouagadougou. A study was performed to compare the technica! and economie aspects of the existing improved Nouna stoves (mason-built, high-mass, multi pan-hole stoves shown in the figures 5.1 and 5.2) and a workshop-made, portable, metal, shielded-fire (figure 5.3). The latter, the Ouaga Metal, was a simplified and redesigned version of Visser's shielded fire

(figure 4.1). Waterboiling tests (VITA procedure) showed the stove's technica! superiority. Moreover, and more importantly. fifteen to twenty times as many Ouaga Metals were constructed by a production unit of comparable size and production costs could be decreased dramatically;

The above results gave new direction to stove projects in Sahel countries. The Ouaga Metal (figure 5.3) was copied; improved in mechanica! strength (Sidibe et al.1983 and Strasfogel 1983); given another name (Mai Sauki, see figure 5.11); and used as the basic design for dissemination in Niamey (Niger). The project was drafted by Krishna Prasad and Bussmann (1983), initiated and supervised by the Energy Department of the World Bank (Floor 1984): and executed by the Cerman

Scope of the mork

Agency for Technica! Cooperation (CTZ) in close collaboration with the Ministry of lndustry and Energy. Technica! back-up services were

supplied by the WSG. The project was very succesful; forty thousand

9

Mai Saukies , twice the projected number, were sold within a time span of twenty months (March 1985- December 1986). The Mai Sauki turned out to be the first stove in a decade, which had the potentlal of taking over the strong market position of the less efficient Malgache stove (figure 5.9). However, the disadvantages of the stove arealso obvious. First. the stove is more expensive than the traditional stove. second. a serious comfort problem is formed by the smoke, and third, the stove has llmited applicability for more than one size pan. In view of this, it was judged necessary to optimize the Mal Sauki design and to come up with a wider variety of stove types.

An

analysis of these kind of design efforts is provided in chapter 5. It

is shown that several levels of designing can be distinguished: the selection of the conceptual designs, the process design, the mechanica! design, the production design, the manufacturing and the marketing (Krishna Prasad 1987). The work involved at all levels is discussed, taking the involvement of the WSG in the stove projects in Burkina Faso and Niger as a case in point.

Most of the actlvities (both in the field and the laboratory) described in this work, were performed simultaneously. Inevitably, this fact has resulted in incompletely developed laboratory prototypes being

introduced in the field. However, there is clearly an evolution towards better designs. This also is the basic concept of chapter 5: designing is a continuous process in which the design campromises reflect both the state of the art in a pure technica! sense, and our knowledge and

11

OIAPTER 2 STOVE TFSTING. EXPERIMENTAL SET-DP AND PROCEDURES

2.1. Introduetion

The cambustion of wood is a complex phenomenon. As a result studies on wood cambustion are still almost entirely empirica!. Emmons (1983) states that: "The current scientific understanding of the cambustion of wood is not mature enough to provide quantitative prediction, but. nevertheless, is very useful for providing direction. The final design of a stove should be determined experimentally for the available bio-fuel guided by what is currently known of the science of fire". Emmons, thus, emphasizes the need for experimental work in designing stoves.

In section 2.2 the test methodology wil! be presented, which is used in the WSG laboratory experiments. The presentation is foliowed with a discussion on the relevant quantities in the sections 2.3 to 2.5. Finally, in section 2.6, these quantities are used in a model which calculates the wood consumption for any given cooking task.

The WSG test procedure is not the only one of lts kind. With the growing interest in improved stoves, many other methodelogies have been

proposed. Most of them focussed on the energy use under actual cooking conditions and failed to give any insight into the relationships among the stove parameters. Consequently they are not of use in designing stoves. A characteristic procedure, the VITA procedure (VITA 1982), wil! be discussed in section 2.9.

2.2. The WSC test procedure

The WSG test procedure for woodburning cookstoves presented in this chapter is basedon the methodology for testing gas ranges (CIVEC 1968). Using a procedure similar to the CIVEC methodology means that:

i. the maximum power of the stove must be determined;

12 Olapter 2

must be measured;

iii. water boiling tests have to be performed, in order to measure the

efficiency of the stove as a function of the power; and

iv. it must be determined whether or not the stove meets the safety

requirements.

The main difference compared with testing gas ranges, is that wood is a solid fuel of non-uniform quality which is not available on tap. The fuel quality is used as a catch-all term to denote the wood species, the as-fired moisture content of the wood, the size of the wood actually employed in the fire etc. The results of a series of experiments will not match when fuel parameters are varled in addition to the variabie actually examined. The consequence is that experiments have to be subjected to a well defined scheme of fuel preparatien and fuel loading. When this scheme is followed, it wil! lead to experimental results which are highly reproducible.

The experimental procedure unless otherwise speelfled is as follows.

White fir wood is cut into pieces of 20N20M67 mm3 _{and dried to constant}

mass in an oven at 105°C (±48 hr). The wood is then divided into lotsof

100 grams each, which are charged to the fire at fixed time intervals. The fire is built on a grate with a diameter of 0.18 m. At the start of the experiments the fire is lit with a propane burner (±30 s). The experiments stop when all the wood bas been burnt. In the experiments

5 kg of water in a pan 0.13 m above the fuelbed is brought to the boil

and is kept boiling. Some water wil! evaporate and escapes as steam. The pan is made from aluminium, bas a diameter of 0.28 m, a height of 0.24 m and is always used with a lid.

2.3. The power output

2.3.1. Nomina! and average power

It is common practice to test heating devices at constant power (burning rate). In the present case this demands a steadily burning fire. One way of obtaining such a fire is by adding smal! quantities of wood at short time intervals, but this is a cumhersome procedure. However, the stove can be considered to burn in a steady periodic way when bigger charges (AMf) are added at larger time intervals (At). For such a batch process

Stove testtng

a nomina! power can be deflned accordlng to the formula: AMfB

At (2. 1)

In general the nomina! power of a fire of a given configuration can be varled within certain well-deflned limits by changlng AMf. At or both.

13

A problem arises due to the build-up of charcoal in the fuelbed. This results in the water continulng to boil even after the last wood charge bas been burned away. Since the end of the water boillng test is defined as the moment of time the water stops boiling (te), lt ralses the question whether it is not better to use a power deflnition based on this time, 'leading to the concept of an average power:

p

av

=

(2.2)

Average powers for correspondlng nomina! powers are shown in table 2.1.

Exp. Parameter Nomina! Average Pav-Pn X

No varled Power kW Power kW Pav

Nomina! Power 1 (with grate) 3.1 2.9 6.9 2 3.9 3.4 H.7 3 5.2 4.5 15.6 4 6.2 5.1 21.6 5 7.8 5.6 39.3 6 (without grate) 2.7 2.7 0 7 3.1 2.8 9.7 8 5.2 4.3 17.3 Moisture content (with grate) 10

ox

3.9 3.4 H.7 11 5% 3.9 3.6 8.3 12 10% 3.9 3.4 14.7 13 15% 3.9 3.5 11.4 H 20% 3.9 3.5 11.4 15 25% 3.9 3.3 18.2 Wood species (with grate) 16 Meranti 3.9 3.4 H.7 17 Beech 3.9 3.5 11.4 18 Oak 3.9 3.5 11.4

Olapter 2

The data presented in the table are from experiments performed on the open fire. The extent of the difference between nominal and average power can be taken as a measure of the charcoal heat available with a given stove configuration. It is a function of power, fuel type, moisture contentand wood block size.In general the difference between nominal and average power increases as the power of the fire increases. The influence of the moisture content and the wood species is erratic; too few experiments were performed to draw any firm conclusions.

Adding the wood in charges creates another unexpected problem. Changing the mass of woodchargedor the charge time interval, in order to vary the power, has completely different effects on the fire behaviour. This is illustrated in the figures 2.1. where the carbon dioxide content in the flue gases is given as a function of time, 2.1a, where the charge weight is varied and the peak values of the carbon dioxide concentration curve change accordingly, and 2.1b where, on the other hand, the charge time is varied but the peak values in the carbon dioxide concentration increase until, during the last charge of the three experiments, they all attain about the same level.

I

20 .-..

1v\

N i \

'"

i

\

0 10 ... +J Cll ~ +J I

'"

Cl) I u

'"

0 u N 0 u 0 0 30 0 30 Time (min)

Fig 2.1: Carbon dioxide concentration versus time for three nominal

powers

a: Charge weight varied: b: Charge time varied:

9.0 kW - - 6 . 3 k W - 3.4 kW 9.5 kW - - 6 . 0 k W - 4.0 kW

Stove testi.ng 15

2.3.2. The time dependent power output

The phenomena shown in the figures 2.1 cannot be explained on the basis of the average and/or nomina! power. There is clearly a need to obtain a better understanding of the time dependent burning process. To that end the experiments are performed on a platform balance in order to record the weight of the fuelbed as a function of time. The balance is shown in figure 2.2: it was designed in such a way that, first, the recorded signa! was insensitive to the position of the fire on the platform: and,

second, the mass of the stoves (upto 60 kg) could be balanced completely

by counterweights. The equipment was insensitive to temperature changes and had an accuracy of one gram (Visser 1982).

Figure: 2.2: Platform balance with acting forces

Every ten seconds the weight of the fuel was measured. This measuring and gatbering of data was done by means of a data acquisition unit (HP 3497A) controlled by a microprocessor (HP 85). After measuring, the data were stored on tape which was processed after the experiments. The set-up of the measuring system is shown schematically in figure 2.3.

The record of the fuelbed weight is a valuable tool for understanding wood fires in operation. In genera!, after an initia! period of unsteady

operation, 'a periodic steady state' regime of burning will be

16 Chapter 2

with one another -of course this will not be achieved in a strict mathematica! sense.

Control Unit

Fig 2.3: Scheme of data processing

A characteristic picture of the fuelbed behaviour of the open fire subjected to the charging procedure is given in figure 2.4a. The figure shows the fuel weight as a function of time. The experimental procedure

is as follows. At time t=t1 a charge AMf is added to the fuelbed, This

charge catches fire and burns until. at time t=t1+ At. a fresh charge

AMf is added, and so on. The refuelling, at t=t1+ At, does not always

mean that the woodchargedat t=t1 bas burned completely. Figure 2.4a

shows, for instance. that the amount of wood left from the third charge is equal to Aa gram. In the graph the accumulated mass of unburned wood just before reeharging becomes larger as time progresses and is equal to Ab gram at the end of the experiment. This process is called the fuelbed build-up.

The weight-loss records can be utilized for estimating the actual time-dependent power output of the fire. The primary assumptions bebind these estimates are that the combustion of wood takes place in two phases (solid phase combustion of charcoal and gaseous phase combustion of volatiles) and that the fuelbed only consists of charcoal and ash. This latter assumption is valid for the entire combustion process at low

Stove testtng 17

nomina! powers and for the steady state period at modest nomina! power levels. For large nomina! power levels however this is not true; the fuelbed will consist of wood in various stages of thermal decomposition.

It would be pure speculation to attempt to identify the proportion of

charcoal and volatiles humt for each charge with the present

experimental technique. However. it is quite inefficient to operatea stove of a given configuration at such high nomina! power levels. A final assumption concerns the charcoal combustion rate, which is taken

200 ~ ~ w ~ ~100 ~ ~ ~ ~ ry ~ ID ~ 0 8 0 0 0 Time(s) 1800 Time(s)

Fig.2.4: Fuelbed weight (a) and power output (b) of the open fire as a function of time

18 Chapter 2

to be constant during the period between two fuel loadings. In reality this may not be the case. However, since the calculated point of time at which the power of volatiles becomes zero synchronizes with the

disappearance of the flames. the assumption mentioned is good as a first approximation (see Appendix I fora more elaborate discussion).

On the basis of the assumptions made, the time-dependent powers P(t),

P (t) and P (t) can be obtained by multiplying the. separated mass flows

V C

with their proper combustion values. A sample result of refashioning the fuelbed weight data in the way described above has been given in

figure 2.4b. The figure shows the rate of heat output of the fire subdivided into the contributions from the charcoal as well as from the volatiles. The lowest, block-shaped, curve represents the rate of heat output of the charcoal, P (t). The lower of the parabola shaped curve

c

represents the rate of heat output of the volatiles, P (t) while the

V

topmost curve shows the total rate of heat output, P(t). The charcoal heat output being block-shaped is a result of the assumed constant charcoal burning rate.

2.3.3. The JDaXimma power

The discussion so far showed the problems in defining the power output.

On top of this, criteria have to be found which make it possible to rate

the power of a stove. The criterion used so far is the excessive

build-up of the fuelbed. The nature of this is shown in figure 2.5 where the fuelbed weight of the open fire on a grate is given as function of time for three different nomina! power levels. In the experiments the varlatlens in power have been accomplished by changing At and holding

AMf constant. On the basis of a change in slope it is possible to

distinguish two reglens in the weight-loss curve of each charge. The point at which a drastic change in slope occurs in general colneldes with the disappearance of flames (the point is marked in figure 2.5a). After this point only the charcoal in wood burns. Figure 2.5c then shows that a fresh charge is added even befere the flames disappear when the nomina! power is raised to 7.8 kW. It leads toa rapid build-up of the fuelbed. Attempts to increase the power beyend that of 7.8 kW result in the fuel falling off the grate. The corresponding situation for an open

Stoue test ing

result in an increased fuelbed diameter. For a closed stove, exceeding the high power limit results in the choking of the cambustion chamber wlth fuel. A glven design of a stove will thus permit a well defined

range of power to be obtained from lt.

o_o----~---L----~1L2 ____ - L ____ _ L _ _ _ _ ~2~4~--__j

Time

(min)----19

Flg.2.5: Fuelbed welght as a function of time for three nomina! powers

20 Olapter 2

Only recently the criterion of the excessively high carbon monoxide to carbon dioxide ratio came into the picture. Many more experiments to collect data in this field are needed.

2.3.4. The minimum power

Criteria also have to be evolved which determine the minimum power or

the turn-down factor. The task to perform at minimum power levels is to balance the convective, radiative and evaporative heat losses from the pan in the siromering period.

The convective heat losses are given by:

q _con

=

A h _{p p} (T - T ) _{p a} (2.3)

The size of the pan area losing convective heat (A). depends on the

p

position of the pan on/in the stove. The heat transfer coefficient for

the pan wall and lid (h ) can be calculated using the Nusselt number

p

relations for free convection (Kreith

&

Black 1980):.

where:

Nu

=

C (Cr Pr)n

C = 0.15 and n = 1/3 for the lid; and

C

=

0.59 and n

=

1/4 for the pan wall

(2.4)

When the cooking is done in a windy environment, the above Nusselt

number relations must be replaced by those for forced convection.

The radiative losses are given by:

q _ra

d

=

~ a

A (T

_{p p} 4 -

T

_a 4) (2.5)

The losses depend on the emissivity of the pan surface which can vary from nearly 0, for bright shining aluminium surfaces, to 1 for black

surfaces (Eckert

&

Drake 1972).

Finally, the evaporative heat losses are given by:

The evaporation rate (m ) depends on the difference between the

eva

Stove testtng 21

saturation pressure at boiling and ambient temperature. Using modified forms of the Clauslus Clapeyron equation (Irvine and Liley 1984), the rate is given by:

kg/s (2.7)

where

c2

=

42.6;

c3

=

-3893;

c.

=

-9.48; and p _a ~ 0

A series of experiments was performed to determine the convective and radiative heat losses from standard pans with a lid (GIVEG 1968). Water in the pan was heated with an immersion heater; and the temperature rise was recorded as a fuction of time and power of the heater. The measured heat losses' from three different sized pans at 97°C are shown in figure 2.6 as solid dots. The experimental data in the figure are compared with calculated curves. In the calculations it is assumed that evaporation can be neglected due to the use of lids. The average

emissivity of the pan surface has been varled parametrically (0, 0.3, 0.7 and 1 respectively).

Emissivity

0 L---~----~~--~----~----~----~

0. I

o.

2 0. 3 0.4

Pan diameter (m)

Fig 2.6: Calculated and measured heat losses versus pan diameter

e

Clean pans 0 Sooty pan

Good agreement between theory and experiment is obtained for an emissivity of 0.3. This value is judged reasonable for the corroded

22 Cha.pter 2

aluminium pans used in the experiments. One experiment. also shown in the figure, has been repeated with a sooted pan. The radiative losses increased considerably; an average emissivity of 0.7 is then needed in order to obtain agreement with the calculations.

The calculated heat losses from standard pans with a lid, having an emissivity of 0.3, receiving heat from the bottorn only, placed in wind-free surroundings, and at boiling temperature, are shown in table 2.2.

Standard Pan Heat Losses Pmin

(71-30%)

Diameter Height Radlation Conveetien Total

(mm) (mm) (Watt) (Watt) (Watt) (Watt)

120 86 9 27 36 120 160 109 15 45 60 200 200 132 23 66 89 295 240 154 33 90 123 410 280 178 44 118 163 545 320 200 57 150 207 690 360 224 72 185 257 855 400 246 89 223 311 1035

Table 2.2: Calculated heat losses from standard pans .

The heat losses shown in the table, divided by the efficiency of the stove, give the minimum powers required to keep the food simmering. In the last column of the table this was done for the improved stoves

disseminated in Niger which have an efficiency of

±

30%. On the basis of

table 2.2. and knoving the maximum power of the stoves in Niger (ranging from 4 kW to 12 kW), it is concluded that the stoves should have a turn-down factor of 10. In reality the turn-down factor ranges from 3 to 4 and consequently the design has to be improved on this part. The designer should aim at a minimum turn-down factor of at least 6; a value specified in the GIVEG standard for gas ranges.

The required minimum power level for open fires and improved stoves without air control, can not be obtained by simply decreasing the charge size and increasing the charge time. In figure 2.7. this is illustrated for the open fire. The figure shows the water temperature response in

Stoue testtng 23

the experiments discussed before (figure 2.5). At a nomina! power of 2.6 kW, the temperature curve exhibits distinct plateaus that correspond roughly to the disappearance of flames. The power level obtained from the charcoal bed apparently is sufficient to balance the heat losses from the pan. Thus for sirnmaring conditions the power should be reduced to this level. However, in the experiments any attempt to lower the nomina! power below 2.6 kW resulted in an insufficient quantity of burning charcoal on the fuelbed for a fresh charge to reignite without the assistance of an external pilot flame.

u <1)

'"'

;::l

...

"'

'"'

<1) ~ <1) E-<

'"'

<1)

...

~ 0 0 Nominal Power (kW)

,

--.,..---"..--~---1 • I I 7.8/

i

5.2 2.6

/

,.

I • I I I •

' /

I • I I I • I I

'

; 30 Time (min)

Fig.2.7: Water temperature as a function of time.

60

The foregoing suggests that improved stoves should be able to separate the volatiles and charcoal combustion, the charcoal being used during the sirnmaring period. In all the (one hour) experiments with the open fire reported here, the water kept boiling after the flames disappeared for a little over 6 minutes at the low power end and 17 minutes at the high power. However, in some of the experimental closed stoves tested by the EindhovenJ'Apeldoorn group, this charcoal heat is sufficient to keep

the water boiling for over 30 minutes (Nievergeld et al. 1981: Vermeer

&

Sleieken 1983). In this way, the charcoal heat can contribute to substantial savings in the fuel consumption.

The heat losses during sirnmaring become much larger when lids are not used and evaporation cannot be neglected. A series of experiments was

24 Cha.pter 2

performed to determine the heat losses from 280 mm diameter pans under

similar conditions as the experiments wlth lids discussed before. A

constant heat flow was supplied to the water with an immersion heater and the resulting steady state temperature was measured. The

experimental data are presented in table 2.3. The steady state

temperature was used to determine the constant C1 of the evaporatlon

rate equation 2.7 .• whereafter the evaporative heat losses at boillng point were then calculated. The table below shows the results for three different power levels of the immersion heater.

Pan Diameter 280 mm

Power immersion heater 400

w

600

w

800

w

Steady state temp. 70°C 82°C go

oe

Calculated heat loss

at boiling point 951

w

991

w

1036

w

Table 2.3: Heat losses at boiling point from 280 mm pan without lid. Calculated on the basis of experiments with an immersion

heater at 400 W, 600 W and 800 W power output.

The results of table 2.3 are consistent. The calculated heat losses only

vary by less than 5%. This justifies the modelling of the evaporative

losses as discussed. Comparison of the tables 2.2 and 2.3 shows that the

heat losses at boiling point can be reduced by a factor of six when

evaporation is prevented, i.e when lids are used and the power is adjusted.

2.3.5. The design power

The power of the fire peaks after fresh wood is added.

A

stove must get

through this power peak with a reasonable combustion quality. A

sufficient amount of combustion air must be supplied. In designing

stoves this has led to the definition of the design power (Pd } which _es

is the maximum nomina! power that a stove can deliver under steady state operation. Experiments on open fires showed that the design power is about 70% of the peak power mentioned above (chapter 3). It is feasible to operate open fires with a nominal power higher than the design power

Stoue testtng 25

for durations of the order of 1/2 to 1 hour. It can be expected that closed stoves exhibit similar tandeneles in their oparation (see for example Delslng (1981) for non-chimney closed stove oparation qualitatively illustratlng fuelbed build-up). However, when closed stoves are operated at powers exceeding the design power on a regular basis, they wil! demand frequent malntenance by way of the removal of tarry deposlts in passages and from the lnternal walls of the stove. When this malntenance is not provided, the performance may deteriorate rapidly and may even lead to drastic reductlons in the lifetlme of stoves. Thus for safe and durable oparation of a stove, it is assentlal to operate it at, or near, lts design power level.

Formula 2.8 finally shows how Pdes can be thought of as being built-up

from the charcoal and volatiles powers. P and P _{c v,max} respectlvely.

Pd _es = 0.1 (P _c + P _v,max ) (2.8)

Insection 2.3.1 it was argued that P is constantduringa charge c

interval. For steady state operations, and assuming a fixed carbon content in wood of 20%, thls implies that:

p = 0.32 p

c n (2.9)

P _v,max on the other hand is the average value of the volatile power

peaks recorded in the weight loss experiments. P wil! play an

v,max

important role in analyslsing the experimental data of chapter 3.

2.4. The efficiency

The efficiency is a maasure of the heat transferred from the stove/fire to the pan(s) under a regime of well-deflned operational procedures. The

efficiency is determined by a water boiling test as discussed in

section 2.2. The efficiency is taken to be the ratio between the heat absorbed by the water in the pan and the heat released by the wood under the assumption of complete combustion. The efficiency is calculated using the formula:

26 Chapter 2

Evaporation is not considered to be a loss, which has led to some

confusion. In the standards of performance for dornestic gas ranges the relevant experiments only last until boiling point is reached

(GIVEG 1968). For wood stoves this would result in large experimental errors due to the fact that the fuel is not available on tap but must be charged in batches. Replacing pans once boiling point is reached might overcome this problem (Micuta 1982}. However, as long as the work is restricted to measuring comparative efficiencies, what metbod is used to calculate those efficiencies is a trivia! matter (Krishna Prasad 1981).

The efficiency depends on the pan size used. If the choice of pan and cooking task is not restricted by the stove design, they are chosen

according to the power rating; thus P needs to be determined first.

max ,-. 00 -~ '-../ .w r:: (]) .w r:: 0 u 1 ;.. (]) .w <ll 2 3 t, 5 6 ;:3: 1 2 1 2 3 4 5 6 7 8 9 10 Power Output (kvJ)

Fig.2.B: Pan diameter and water content as a function of the power

output for three different power densities at the pan bottom.

Upper sca1e 7W/cm2, middle scale 10 W/cm2, and lower scale

Stove testtng 27

The rule of thumb for gas ranges is to have a heat flux through the pan

bottorn of 3.5 W/cm2• At an efficiency of 50% this requires a power

density at the pan bottorn of 7 W/cm2• For wood stoves the power density

should be much higher as the efficiency is lower. Thus the x-axls in flgure 2.8 bas three different scales, reprasenting power densities of

7 W/cm2, 10 W/cm2 and 17.5 W/cm2 bottorn area at efflciencles of 50%, 35%

and 20% respectively.

2.5. The

co-co?

ratio

One of the major problems in designing lmproved stoves concern the combustion quallty. In testing woodburnlng cookstoves this problem has almost been neglected entirely although Smith (1986) clearly showed the seriousness of the problem. The products of incomplete combustion are carbon monoxide, hydrocarbons and soot. Carbon monoxide especially deserves full attention when considering the stove performance. Sulilatu (1985) showed the polsonousness of this colourless and odourless gas. The effect of the carbon monoxide concentration in the atmosphere as a function of the exposure time for varlous conditlens of labour are shown in figure 2.9.

~ 0 0 LJ 0.2 0.1 0.01

""'

""-Safe ... r--1---" 10

Exposure time iminl

-...

oladiJlll=l

-

rf

-Poisoning ... I '

-

r-1111

100 1000 10000

Fig.2.9: Toxicity of carbon monoxide as a functlon of the exposure time

Sangen (1983) showed that the carbon monoxide content in the flue gases also correspond to the release of unburnt hydrocarbons (see

figure 2.10). This is also impliclt in the more recentworkof

28 Chapter 2

C H and soot, the carbon monoxide content has been taken as a quantity x y

indicative of the combustion quality. In spite of this simplification, the experimental task of measuring the quantity of carbon monoxide emitted in non-chimney stoves is hard.

QS

•

~

0 V 0 3000 4000 0 1000 2000 CxHtppml

Fig.2.10: Relationship between carbon monoxide and C H concentrations. x y

Absolute measurements of the carbon monoxide concentration in flue gases are only possible in closed stoves with a chimney. This is why for all practical purposes, the carbon monoxide to carbon dioxide ratio is used as an indicator of the toxicity of the combustion gases. It involves the measurement of both the carbon monoxide and carbon dioxide content. The expertmental set-up is shown in figure 2.11. (Sielcken 1983).

CHIMNEY

Stove testing 29

The methodology for testing gas ranges gives clear standards for the relative carbon monoxide content in the flue gases. In table 2.4 the norms applied in the Netherlands for different humers are glven. It is essential that the power is speelfled as well.

Stove type Power

co-co2

ratio

gas appliances Pmax

<

1.0%

kerosene burners Pmax

<

1.2%

Anthracite burners Pmax

<

2.0%

Dornestic space heaters

using wood Pmax

<

4.5%

Pmax/2

<

9.0%

Table 2.4: Carbon monoxide to carbon dioxide standards for wood stoves

r

N 0

u

9.0 _ Dornest ie. space. heaters (Pmax/ 2:) __ --j

----.: I

~~

!

Gombustion chamber diameter./

1 x 20 cm 1 o 25 cm I o 30 cm /

',~

a::::·~

... ~ ... ·~

...

·

,

"

p

I ( ... 0'

-- --c--.",..

---Dornestic space heaters (Pmax)

0

1 2.0 .Anthracite_burners_(Pmax)_ __

8

1. 2

1.0

_Kerosene_burners _(Pmax) _____ _ -ca s- app liane es-

(Pmax)----3 5

Power Output (kW)

7

...

Fig.2.12: Carbon monoxide to carbon dioxide ratio as a function of the nomina! power. Data from stoves with combustion chambers of different diameter

30 Chapter 2

In figure 2.12 the carbon monoxide to carbon dioxide ratio is shown as a function of the power for three simllar chimney stoves, with combustion chambers of different size. The carbon monoxide to carbon dioxide ratlos were averaged over the water boiling experiment. In the figure the

standards from table 2.4 are shown too. Although the stoves show higher

efficiencies than many other designs, their index of toxicity is dangerously high for the whole power range investigated. lt shows the necessity for including the carbon monoxide and carbon dioxide

measurements in the test methodology.

2.6. The specific consumption

Once data on the maximum power, the minimum power and the efficiency are

available. it becomes possible to calculate the time required for

cooking and the specific consumption (SC) (defined as the ratio between

the mass of wood used and the mass of food cooked) for any given meal

(Krishna Prasad et al. 1983). For this, the cooking task must be

modelled and thus is done through using the water equivalent of the food

cooked and the simmer time (Verhaart 1983). The water equivalent is

defined as: n

2

Mf j Cp.

M

...

~o=·~-

...

~J~ w = Cp (2. 11) J=l

It is assumed that steam production is an inevitable, albeit useless, part of the siromering process. Thus the initia! quantity of water is

larger than the final quantlty,

M .

_w,e The extra water is given by:

M

eva

The heat needed to bring the food to the boil is then:

and the time to boil:

Cp

AT

TJmax

(2.12)

Stove testlng 31

(2.14)

The importance of tb lies in the fact that a long cooking time is

unacceptable from the user's point of view. The energy used during the

simmering period t is equal to:

s

(2. 15) The specific consumption (the ratio between the mass of the quantity of fuel used and the food cooked) and the total cooking time are finally given by: SC

₌

_M--B

1 l w,e t e Cp AT + p t

l

Tl _max min s (2.16) (2.17)

The above formulae can he simplified. Cp and

L

are known constants; AT

is normally ±75°C. In addition there is sufficient evidence available to suggest that the efficiency of a stove is not a strong function of the

power and the mass of food cooked. Thus as a first approximation Tl _{ma x}

equals Tlmin· Using the concept of the turn down factor the formulae for

the SC and the boiling time finally become:

SC _1_ [ 315 +

_B

~

Tl 1.14 _e (2.18) M [ p t and: t _e = w,e p ~+0.14 _{Tl M} s _r + t _s w,e (2. 19)

The equations show that a reduction in SC can he achieved by: increasing

the as-fired combustion value

B;

increasing the efficiency TJ; decreasing

the power rating of the stove P; increasing the mass of food cooked

M _w,e ; reducing the simmering time t _s and increasing the turn down

32 Chapter 2

2. 7. The VITA test procedure

The charging procedure outlined in the sections 2.2 to 2.6 has evoked a lot of discussion. Alternative procedures have been proposed which claim to reflect reality much better. The philosophy underlying these

procedures is that testing a device is carried out to assess its

performance in relation to the task it is expected to do. The VITA test methodology follows such procedures; it was meant to be an international standard for testing woodburning stoves (VITA 1982). The methodology has three levels of testing. The first level is a simple simulation, with water, of a standard cooking procedure. The results are expressed as the standard specific consumption. SSC, which is the ratio between the water vapourized and the wood used. The second level is the controlled cooking test which involves the cooking of a selected meal. The results are expressed as the specific consumption, SC (the ratio between the wood used and food cooked). The third level is the kitchen performance test whlch measures the relativa rate of fuel consumed by two stoves when they are used in the normal household environment. The results are expressed as the specific daily consumption. SDC, which is the ratio between the wood use per day and the family size. The following are problems with the VITA procedure.

i. The definitions of the SSC, the SC and the SDC do not relate with

each other. Without additional information it is impossible to use the results from one test level at other levels. The three test levels might even show contradictory results. This occurs. for instance, when there is a large difference in the duration of the water boiling test and the controlled cooking tests.

11. The procedure speelfles only how to performa test but does not

specify how to evaluate the results. No standards are given against which the performance of the stoves can be assessed.

iii. Safety aspects are completely ignored. Additional restrictions have to be placed on the equipment since it has to be operated by a human being. These restrictions must come in the form of safety to the operator. It is customary to demand stringent safety

requirements from household appliances since these are operated by a large and diverse population.

iv. The procedure is not sufficiently clear to enable an experimenter to obtain reproduelbie results. It gives an Imprassion of being

Stove testtng 33

very accurate while, in fact, the reproducibility of results is poor. This is shown in figure 2.13 where the standard deviation of the boiling times has been plotted. In the figure, data are shown

from IVE (1983), Wouters (1984) and Strasfogel

&

Dechambra (1984).

The IVE data cover 130 boiling tests performed on 13 stoves.

Wouters tested four stoves and Strasfogel two, covering in total 20 and 28 experiments respectively. The standard deviations in the boiling times measured with the WSC test procedure are also shown in the figure. The latter standard deviations are much smaller (which was the desired result).

1

50 40 ,-.. 30 ll'.!; .._" c: 0 ·rl 20 ..., Cl! ·rl > Q) 0 '0 10 1-< Cl! '0 c: Cl! ..., 0 Cf) 0 10 Average

•

....

•

•

_•

x••

....

".

•

x.

....

0 0 0 20 30 time (min)

•

•

40

•

Fig.2.13: Standard deviations in measured boiling times

• IVE, A Wouters, X Strasfogel, 0 WSC

The points mentioned show the VITA methodology to be ill-suited for use in design work.

2.8. Concluding ramarks

The advantages of the WSC procedure are summaried below.

i It gives detailed information on the basic characteristics of a

stove (efficiency, power range and cambustion quality), which play a key role in the design work presented in chapter 5.

34

ii. On the basis of the power output/efficiency concept, it becomes

possible to model the energy requirements for every cooking task. iii The experiments give highly reproduelbie results (figure 2.13),

which is probably the most practical advantage.

The discussion also revealed some difficulties. The understanding of processas in woodburning cookstoves is still at such a level, that a clear and unique definition of the power output is difficult to give (section 2.3). Experimental work is still required tobetter relate the

different power concepts proposed. A last point to mention is that, so

far, no data whatsoever has been reported on safety aspects like

stability, wall temperature and combustion gas leakage. It is believed