Stove Design & Performance
Testing Workshop
WHO IAP Workshop
Kampala , Uganda
June 17
th, 2005
Presented by Peter Scott ( [email protected]) With special thanks to Mike Hatfield and Rob Bailis
With protocols developed by Rob Bailis, Dean Still, and Damon Ogle with input from Dr. Kirk Smith and Dr. Rufus Edwards
The University of
California at Berkeley’s Energy and Resources Group and School of Public Health
Simplified stove theory
• Wood doesn’t burn
• Wood gets hot and releases volatile gases that then
combust
• For this to happen we need to have sufficient time
temperature and turbulence
• If wood is heated to 650 degrees Celsius (and sufficient
oxygen is mixed with the volatile gases) the result is
complete combustion . The products of clean combustion
are CO2 , water vapour and heat.
• A lot of heat , roughly speaking , dry wood has half the
energy per kg as gasoline,
What are limiting factors to high temperatures ?
Challenge # 1
• Cool stove body
• Cool earth
• the body of the stove or of
the earth robs heat from the
fire
• which lowers combustion
temperatures… which
decreases efficiency…and
increases smoke
What are limiting factors to high temperatures ?
Challenge # 1
• Cool stove body
• Cool earth
• the body of the stove or of
the earth robs heat from the
fire
• which lowers combustion
temperatures… which
decreases efficiency…and
increases smoke
Solution?
• Insulate the stove
with low mass, heat resistant materials in order to keep the fire as hot as possible• Remember mass is the opposite of insulation
• Effective stove insulators are pumice , vermiculite, and wood ash
• Dense things such as earth,sand, cement, water and cast iron are poor insulators
Maximizing combustion efficiency
• Challenge
#2
• Cool wood
• which lowers combustion
temperatures…which decreases efficiency…And increases
Maximizing combustion efficiency
• Challenge
#2
• Cool wood
• which lowers combustion
temperatures…which decreases efficiency…And increases
smoke
• Solution?
• Meter the fuel!
• Use small sticks whenever possible • Maximize the surface area of the
wood exposed to coals
• Heat only the fuel that is burning • Burn the tips of sticks only as they
Maximizing combustion efficiency
• Challenge # 3
• Cool air/ Too much air
• which lowers combustion temperatures… which
decreases efficiency…And increases smoke
• Note: an open fire can draw 20 times more than is required for stochiometric (chemically ideal) combustion
Maximizing combustion efficiency
• Challenge # 3
• Cool air/ Too much air
• which lowers combustion temperatures… which
decreases efficiency…And increases smoke
• Note: an open fire can draw 20 times more than is required for stochiometric (chemically ideal) combustion
• Solution ?
• Do not allow too much or too little air to enter the combustion chamber.
• there should be a
minimum excess of air supporting clean burning.
Maximizing combustion efficiency
• Challenge # 4
• Cool cooking pot
• The cooking pot is
generally no more
than a 100 –200
degrees Celsius
• Flames touching the
pot?
Maximizing combustion efficiency
• Challenge # 4
• Cool cooking pot
• The cooking pot is
generally no more
than a 100 –200
degrees Celsius
• Flames touching the
pot?
• Soot and smoke!
Solution?
w Elevate the pot above the
height of the flames
• This creates an internal
‘chimney’ which increases
draft
• And gives time for
Optimising heat transfer
• Maximize surface area
of pot that is exposed to
hot flue gases
• Maximize velocity of hot
flue gases to disturb
boundary layer
• Maximize Delta T.
Temp difference between
hot gases and pot. Not all
fires are the same
temperature
• Exit temperatures not to
exceed 180C
With a heat exchanger,
overall efficiency can be
improved by 50% or more
Rocket stove heat exchanger/skirt
• Minimize the gap between
the skirt and the pot while
maintaining the cross
sectional area of the
combustion chamber ( for
average size pots 1cm is
good rule of thumb)
• Make it as tall as feasibly
possible
• One of the keys to producing a smokeless Rocket stove is to find inexpensive, local, and durable materials for the combustion
chamber. In Malawi, we have been blessed to work with Dedza Pottery. They have helped us produce an
insulative refractory brick from sawdust grog and clay that is light (0.67 g/cc) and durable.
•
• In Lesotho we are using cement fondue with ground and graded waste pumice.
•
• In Uganda we are using cut pumice stone.
This open fires use 170 kg of wood to cook corn porridge for 110 people (approx 80.5 kg of cooked food).
A visual comparison between the quantity of wood used
(170kg) for the open fire vs. the amount of wood used (13kg)
by the 100L Rocket stove. Independently tested by EP
200L Rocket Stove
• uses 9.5 kg of wood to
cook enough Nsima
(corn porridge) for 225
people;
• Approximately 160 kg
less wood to cook twice
as much food.
A few rocket stove design possibilities
Can chimneys increase IAP?
• Chimneys burn out or get clogged
This will decrease efficiency and eventually
will increase IAP.
• If we have poor combustion with a chimney
are we just exporting the smoke to our
neighbors?
Rocket 100-300L (Uganda)
• 100Litre WBT
• PHU Efficiency
• 49% without chimney
• Boiled 75 Litres of
water in 52 min (no
lid) approx 6 kg of
wood
100Litre WBT
36% with chimney
X
1.5-2X
TH= X + 1.5X + 5 cmBasic Rocket
Stove Geometry
2X
X
More household Rocket prototypes
• In Kenya, a number of
modified rocket stoves were made with a shorter internal chimney. More smoke, but guaranteed heat transfer with a pot that fits directly into the stove body
• In Malawi , it also comes in a 100% ceramic version for the home.
Before implementing a improved cook
stove project…
Questions to ask
Are people cooking outside or inside
?
Inside
Leave them there! Outside.
Leave them there!
Prioritize Heat Transfer Eff . C Eff is less important
Before implementing a improved cook
stove project…
Questions to ask
Are they presently using biomass fuel for cooking?
If no or have other fuel options
Don’t promote biomass ICS!
If yes
Are they paying for fuel?
If yes consider commercial
approach
Consider User priority e.g. Refugees
fuel saving? Smokeless? Clean pots? Social status? If no or are very low
income
Disseminate via teaching circles
Rationale for the SPT
Demonstrate impact of ICS projects using methods
that are…
– Standardized and repeatable
– Comparable within and across projects
– Statistically sound
…but still appropriate and flexible enough to adapt
to local circumstances and constraints!
– Caveat: Monitoring is important but question of
allocation of resources
And because everything…
Stove Performance Testing (SPT)
Past and present
• In 1985 Vita developed a set of protocols for testing stove performance
• Functional yet somewhat cumbersome . Not generally used
• In 2003 Shell/EPA request UC Berkeley and Aprovecho to develop a new set of
universally adopted SPT protocols
What is Stove Performance?
Measures of Stove Performance
1.
Efficiency/exit temp
2.
Fuel consumption
3.
Turn-down ratio (TDR)
4.
Speed of cooking
5.
User satisfaction
6.
Emissions (?)
Efficiency
Combustion efficiency Net stove efficiency (PHU) Heat transfer efficiency Fuel consumption Speed of cookingUser
satisfaction
Ease of use Durability Flexibility Aesthetic appeal Turn-down
1. Efficiency Entirely lab-based
• Combustion efficiency
- Difficult to measure directly
- Can be approximated by measuring PICs (e.g. Smith, Uma et al. 2000) • Heat transfer efficiency
- Very difficult to measure directly • PHU ( Percentage Heat Utilized)
-Overall
Efficiency
Combustion efficiency CE Overall efficiency (PHU = CE x HTE) Heat transfer efficiency (HTE)Measures of performance
Combustion efficiencies and PHU for 28 stove-fuel combinations in India (each data point is the mean of three tests)
0.70 0.75 0.80 0.85 0.90 0.95 1.00 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70
Net stove efficiency (PHU)
N o m ina l C o m bus ti on E ff ic ie n c y
Liquid and gaseous fuels Solid biomass fuels
Charcoal Charcoal
1. PHU
- Fairly easy to measure directly (assuming fuel HV is known)
-Energy into the food Fuel energy
consumed
Measures of performance
• It is measured by the change in the temperature of the water and
the quantity of water that is evaporated
• So if we compare 2 stoves, the one with the higher PHU is the
‘better’ stove right ?
•Not necessarily! Because PHU rewards the stove that makes a lot of
steam
Specific Consumption vs. PHU
Stove 1
• Time to boil 10 minutes • Wood burned 1000 grams • Water vaporized 100 grams • Water remaining 4.9 liters
Stove 2
•Time to boil 100 minutes •Wood burned 1000 grams •Water vaporized 1000 grams •Water remaining 4.0 liters
Specific Consumption vs. PHU
Stove 1
• Time to boil 10 minutes • Wood burned 1000 grams • Water vaporized 100 grams • Water remaining 4.9 liters
Stove 2
•Time to boil 100 minutes •Wood burned 1000 grams •Water vaporized 1000 grams •Water remaining 4.0 liters
•
•
2. Fuel Consumption
– Lab-based (WBT)
- Very easy to measure
- Multiple definitions possible – Field-based (KPT)
- Ranges from easy to difficult to measure depending on study design and desired outcome.
3. Turn-down ratio (TDR) or Control efficiency
– Ratio of high to low power
• Stoves with large TDRs can be operated more efficiently and may be more preferable to users
4. Speed of cooking
Can be either lab or field-based – Lab-based, easy to measure
- Cooking is simulated (not directly predictive of real household use) – Field-based
- Can be measured directly, but better to rely on surveys
5. User satisfaction
– Hard to measure, subjective, and dependent on many factors
Fuel consumption Speed of cooking
User
satisfaction
Ease of use Durability Flexibility Aesthetic appeal
Water Boiling Test
1 -
Based largely on VITA (1985) and Baldwin (1986) with small modifications- Limits Variables
- Transferable between various projects
Lab-based test provides 4 of the 5
indicators of SP:
1. PHU
2. Specific Consumption 3. TDR
4. Time to boil
-But its difficult to extrapolate these results to actual field performance
Overview of the WBT
Modifications to VITA’s WBT:
– Specific Consumption defined as the ratio of the total amount of wood used to the amount of water boiled, but was modified for multi-pot stoves.
– Low power test is done as a separate test, which allows for a more relaxed procedure than VITA’s test with minimal loss in accuracy.
– For the low-power test, the tester keeps the water temperature as close as possible to 3°C below boiling in order to reduce variation in test results (earlier tests used a 5 degree range).
– Hot and cold starts are used in the high power phase to account for differential performance of stoves that are kept hot throughout the day (important for massive stoves with performance that varies between cold and hot starting conditions).
– Simmering time in low-power test is longer than VITA, but shorter than Baldwin (45 minutes rather than 30 or 60) . This is a compromise: long enough for the stove at low power to establish equilibrium but reduce the time required for the test.
Controlled Cooking Test
(CCT)
- lab controlled test with
added variables of an actual cook cooking real food
- Only can be used to
compare two stoves from a particular
- Compares fuel consumption (specific consumption), and speed of cooking
- Much better at predicting actual stove performance and fuel consumption in the field
Kitchen Performance Test
KPT
-More complex than WBT:
–Both qualitative survey and quantitative measurements
–Takes stove testers into peoples households
–Sampling procedure and study design are critical
–Variability in “real-world” setting
increases the number of samples needed to make results statistically valid (more later).
–Gives daily wood consumption and gauges user satisfaction
Overview of the WBT
Each WBT consists of 3 parts:
–
High-power cold start
–
High-power hot start
–
Low-power (simmer)
And takes roughly 2 hours to complete
We recommend 3 tests of each type of stove
Sufficient to detect a 30% improvement with 95% confidence if the pooled CV of measurements is 15% and a 20% improvement in PHU if the pooled CV is
Overview of the KPT
Also based largely on VITA (1985) with inputs from
FAO (1983) and others.
Field-based test provides two indicators of SP:
1. Fuel consumption
2. User satisfaction
Based on actual measurements from households using
the ICS.
Principal procedure for Shell HEH projects to demonstrate that
they have met their stove performance objectives.
Review of KPT procedures
Qualitative surveys
– Include two surveys
• Pretreatment – designed to assess situation in
households before use of improved stoves
• Post-treatment – designed to assess qualitative
impact of stove on household
– If group already has surveys that they use,
these should provide additional ideas for
questions
Review of KPT procedures
What is included in qualitative surveys?
Preliminary survey:
–
Household characteristics
–
Current cooking patterns
• Who cooks?
• What type(s) of stove and fuel are used? • Who collects fuel?
• How much fuel is used?
–
Is the family interested in receiving a new stove and/or
participating in future surveys
Review of KPT procedures
What is included in qualitative surveys?
Post-treatment survey:
– Any changes in household characteristics since
receiving the stove
– Use of the new stove and other stoves
– Stove maintenance
– Stove performance and user satisfaction
Review of KPT procedures
Sample selection for qualitative surveys
Issues to consider:
• Clustering
• Randomization
• Number of households
100 Large (more than 1000 households)
~10% Medium (300-1000)
At least 30 Small (less than 300 households)
Number of households to be surveyed
Size of community ( or group of communities)
Review of KPT procedures
The quantitative field test
• Main output is average fuel consumption per person • Sample selection
– Similar issues as in qualitative survey but quantitative nature of data makes study design very critical.
– Clustering?
• Are there qualities in different communities that affect fuel consumption (climate, wealth, fuel availability, etc.)?
If yes, communities should be treated as different populations and tested separately!
– Randomization?
• Should be done if possible – important to generalize results of tests. • Requires census of households within target communities
Review of KPT procedures
Number of households
– Must decide on cross-sectional or paired
sample design
Cross-sectional – different groups of households
are tested using TCS and ICS.
Paired sample – the same group of households is
tested first using TCS and later using ICS
Review of KPT procedures
Deciding on the number of households:
–
The variability of the data (how scattered is it?)
–
The difference that you want to detect in average fuel
consumption between TCS and ICS
Review of KPT procedures
Deciding on the number of households:
Hypothetical results (in kg-wood per person-day)
of 3 sets of measurements
58% 31% 20% CV = 0.8 0.4 0.4 Std Dev = 1.42 1.40 2.00 Mean = 2.1 1.1 1.6 Household 3 1.7 1.9 2.4 Household 2 0.5 1.2 2.0 Household 1 Improved stove 2 Improved stove 1 Traditional stoveReview of KPT procedures
(sample) 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0 1 2 3 4Traditional stove Improved stove 1
0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0 1 2 3 4
Traditional stove Improved stove 2
28 14
Sample size (cross-sectional test)
14 7
Sample size (paired test)
-30% -30% % difference in means 40% 29% Pooled CV = 0.7 0.5 Pooled Std Dev = 1.7 1.7 Pooled Mean = TS and IS-2 TS and IS-1 Pooled data
Wrap-up of SPT Training
Results of our trial WBT
Charcoal 2.26 3.6 165 38 Rocket 1.42 13.8 680 10 Open fire TDR Firepower
Specific fuel consumption Thermal efficiency SIMMER Charcoal 8.3 88 43 Rocket stove 19.7 266 9 Open fire Firepower Temp-corr sp consumption Thermal efficiency HOT START Charcoal 5.3 94 24 Rocket Stove 19.7 266 9 Open fire Firepower kW Temp-corr sp consumption grams/litre Thermal efficiency PHU% COLD START
Wrap-up of SPT Training
Results of our trial WBT
Comparison of Thermal efficiency
-0.05 0.10 0.15 0.20 0.25 0.30 0.35 0 .45
open fire rocket Charcoal COLD START
HOT START SIMMER
Comparison of Specific consumption
-100 200 300 400 500 600 700 800 900
Open fire Rocket charcoal COLD START
HOT START SIMMER
Comparison of Time to boil
-10 20 30 40 50 60
open fire SuROcket charcoal COLD START HOT START
