An in Depth Exploration of Cooking Entirely With Electricity in Kenya

An In-Depth Exploration
of Cooking Entirely with

Electricity in Kenya
i
AN IN-DEPTH EXPLORATION OF COOKING ENTIRELY WITH ELECTRICITY IN KENYA
Document title: In-depth Exploration of Cooking Entirely with Electricity

Modern Energy Cooking Services (MECS)
13 February 2023
‘This material has been funded by UKaid from the UK government and is implemented by
the Foreign, Commonwealth and Development Office; however, the views expressed do not
necessarily reflect the UK government’s official policies.’
DISCLAIMER
While this document has been prepared in good faith and based on international best
practices in research and consulting, EED Advisory does not accept responsibility for any
consequences of its use. Readers are responsible for assessing the relevance and accuracy
of the content of this publication. EED Advisory will not be liable for any loss, damage, cost
or expense incurred or arising because of any person using or relying on information in this
publication.
ACKNOWLEDGEMENTS
EED Advisory is pleased with the successful completion of the In-Depth Exploration of
Cooking Entirely with Electricity in Kenya. This study was commissioned by Modern Electric
Cooking Services (MECS).
This report would not have been possible without the collective support provided by
several individuals and organizations. This includes EED Advisory team members – Alois
Mbutura (Team Lead), Ruth Gichuhi, Murefu Barasa, Ann Kahihia, Poulman Aineah, and
Festus Kipkorir, Lewis Waswa and Sylvia Manono. In addition, we would like to appreciate
Phoebe Oriama and Risper Wawira, and the Burn Manufacturing team for their assistance in
procuring electric cooking appliances at short notice. Tash Perros and Arthur Gohole from
Clean Air Africa and the Kenya Medical Research Institute respectively, for their assistance
with stove temperature monitoring tools. Finally, we would like to appreciate MECS team
members, in particular Louise Medland, and Mani Thompson for their support in reviewing
this report
© EED Advisory Limited 2023. All rights reserved
EED Advisory Limited

90 James Gichuru Road, Ikigai Lavington, Nairobi – Kenya
P.O. Box 66053-00800, Nairobi, | T: +254 (20) 2574927
E: [email protected] | W: www.eedadvisory.com
ii
Contents
List of Figures iv
List of Tables v
List of Acronyms vi
Terminologies vii
Executive Summary ix
01. Introduction 1
1.1 Background 1
1.2 Purpose of the study 3
02. Approach and Methodology 4

2.1 Assessment of the available electric cooking devices 4
2.2 Data collection 7
2.3 Implementation of the surveys 13
2.4 Synthesis and Reporting 14
03. Analysis and results 15

3.1 Household characteristics 15
3.2 Cooking diaries overview 17
3.3 Time, energy, and cost savings 36
04. End-user perceptions and barriers to

cooking entirely with electricity 44
4.1 Assessment of e-cooking experience indicators 45
4.2 User perceptions of electricity compared to baseline primary fuel 46
4.3 User Perceptions and Suggestions on e-cook appliance features 47
4.4 Addressing barriers to cooking entirely with electricity 47
05. Conclusions and Recommendations 48

5.1 Conclusions 48
5.2 Recommendations 49
Annexes 51
Annex 1: Cooking Diary Form 51
Annex 2: Entry Survey 52
Annex 3: Exit Survey 55
List of Figures
Figure 1: Summary of Approach 4 Figure 22: Percentage usage of electricity 26

by dish in the baseline
Figure 2: A bar graph of the most common 5
dishes cooked by the 10 selected household Figure 23: Percentage usage of electricity 27
by dish in the transition
Figure 3: Contextual map of the study 8
area, Kawangware, in Dagoretti North Figure 24: Percentage usage of charcoal by 28
Constituency in Nairobi City County dish in the baseline
Figure 4: Mechanical support for LPG, 9 Figure 25: Percentage usage of charcoal by 28
kerosene, and firewood measurements dish in the transition
Figure 5: The Sonoff app 10 Figure 26: Most frequently cooked dishes 29
10 on the gas stove (6-kg) in the baseline phase
Figure 6: Data flow of minute-by-minute
electrical measurement data Figure 27: Most frequently cooked dishes on 30
the gas stove (6 kg) in the transition phase
Figure 7: Sonoff POW Elite monitoring an 11
EPC Figure 28: Most frequently cooked dishes 30
on the M-gas double burner gas stove in the
Figure 8: EPC energy usage within the 11
baseline phase
transition phase on the Google Colab toolset
Figure 29: Most frequently cooked dishes 31
Figure 9: The main data collection steps of 13
on the M-gas double burner gas stove in the
the study
transition phase
Figure 10: E-cooking training using the EPC 14
and hotplate
on the charcoal stove in the baseline phase
Figure 11: Baseline stoves owned across 15
Figure 31: Most frequently cooked dishes on 32
households
the charcoal stove in the transition phase
Figure 12:Mean number of people catered 18
Figure 32: Most frequently reheated dishes 33
per meal classified by study phase
on the microwave in the baseline phase
Figure 13: Average trends in meal 19 Figure 33: Most frequently reheated dishes 33
complexity by the number of dishes
on the microwave in the transition phase
prepared per meal
Figure 14: Complexity of breakfast meals by 20
on the EPC in the transition phase
the number of dishes prepared per meal
Figure 15: Complexity of lunch meals by the 21
on the air fryer in the baseline phase
number of dishes prepared per meal
Figure 16: Complexity of supper meals by 21
on the air fryer in the transition phase
the number of dishes prepared per meal
Figure 37: Most frequently cooked dishes on 35
Figure 17: Changes in dish state during 23
the electric hotplate in the transition phase
supper meals across baseline and transition
phases Figure 38: Baseline energy per capita 41
required per dish against the level of
Figure 18: Changes in dish state during 23
transition
lunch meals across baseline and transition
phases Figure 39: Transition energy per capita 42
required per dish against the level of
Figure 19: Changes in dish state during 24 transition
breakfast meals across baseline and
transition phases Figure 40: Baseline cost per capita per dish 42
against the level of transition
Figure 20: Percentage usage of LPG by dish 25
in the baseline Figure 41: Transition cost per capita per dish 33
against the level of transition
Figure 21: Percentage usage of LPG by dish 26
in the transition
iv
List of Tables
Table 1: Summary of barriers xii Table 16: Cooking time and energy used in 36
cooking fresh dishes
Table 2: Frequency of reasons for not xiii
cooking with electricity Table 17: Cooking time for various fresh 37
dishes across cooking devices in minutes
Table 3: Use rates of electric appliances 1
under the Kenya Cooking Sector Study Table 18: Fresh dish energy per capita 37
(MoE,2019) requirements for cooking across various
cooking devices in Megajoules
Table 4: E-cook appliances chosen to be 6
provided Table 19: Cooking time for various 38
reheated dishes across cooking fuels in
Table 5: Summary of stoves owned across 16
minutes
all households as recorded in the entry
survey Table 20: Cooking time for various 38
16 reheated dishes across cooking devices in
Table 6: Frequency of use of cooking
minutes.
devices within the baseline phase
Table 21: Reheated dish energy per capita 38
Table 7: Frequency of use of cooking 17
requirements for cooking across various
devices within the transition phase
fuels in Megajoules
Table 8: Classifications of households by 17
Table 22: Mean cook time per meal 39
income group
Table 23: Per capita energy requirement 39
Table 9: Mean number of adults and 17
statistics per meal across baseline and
children cooked catered per dish across
transition
research phases
Table 24: Baseline energy per capita per 39
Table 10: Number of times each meal was 18
day
cooked per research phase
Table 25: Transition energy per capita per 40
Table 11: Mean number of adults and 22
day
children cooked catered per dish across
research phases Table 26: Summary statistics on energy 40
and cost dynamics of the baseline phase
Table 12: Fuel use by percentage during 24
(using all baseline fuels – electricity, LPG,
supper meals across baseline and transition
and charcoal)
phases
Table 27: Summary statistics on energy 41
and cost dynamics of the transition phase
lunch meals across baseline and transition
(using all transition fuels – electricity, LPG,
phases
and charcoal)
Table 28: Summary of barriers 44
breakfast meals across baseline and
transition phase Table 29: Frequency of barriers 45
encountered
Table 15: Overall usage of cooking devices 29
across dishes
v
List of Acronyms
SDG7 Sustainable Development Goal number 7

MECS Modern Energy Cooking Services
W Watts
kWh kilowatt-hour
L Liters
LPG Liquified Petroleum Gas
SE4ALL Sustainable Energy for All
KES Kenya Shillings
USD United States Dollar
M Million
NACOSTI National Commission for Science, Technology, and Innovation
MS Microsoft
LCD Liquid Crystal Display
SIM Subscriber Identity Module
FGD Focus Group Discussion
HH Household
AM Ante Meridiem
PM Post Meridiem
MJ Megajoules
ECA Electrical Cooking Appliances
SSID Service Set Identifier
API Application Programming Interface
USA United States of America
S3 Simple Storage Service
TB Terabyte
KNBS Kenya National Bureau of Standards
EPRA Energy and Petroleum Regulatory Authority
KPLC Kenya Power and Lighting Company
vi
Terminologies
Dish: A unit of a cooking activity that can be cooked with a single cooking utensil and requires a
single heating event, e.g., rice
Meal: A combination of dishes prepared and served jointly, e.g. rice and beef stew.
Cooking solution: Any combination of a cooking stove and the fuel used for cooking.
Primary cooking solution: the cooking solution that is most used (frequency of use).
Stacking of cooking appliances or fuel: a household using multiple cooking solutions or fuels to
meet their cooking needs
The exchange rate at the time of reporting 1 US$ – KES 121
LPG market rate per kg – 1.84 US$
Electricity market rate per kWh – 0.176 US$
Charcoal market rate per kg – 0.743 US$
The calorific value of charcoal – 29.9 MJ/kg
The calorific value of LPG – 44.8 MJ/kg
The conversion efficiency of electricity – 3.6 MJ/kWh
vii
viii
©Unsplash
Executive Summary
Introduction
The triggers, enablers, and causes of transitions The framework further allows comparing the
in cooking fuels at the household level have been standard cooking scenario (without e-cook inter-
studied extensively over the last three decades. A vention) and a 100% e-cook scenario by splitting
particular area of study has been the transition of a study into two phases, an initial baseline phase
low and middle-income economies from biomass with regular cooking using standard fuels and a
to electric cooking. In these studies, the low uptake subsequent transition phase with the introduced
of cooking with electricity has been attributed to electric appliances. The baseline phase in this study
the high up-front cost of electric stoves compared was performed in December 2022, coinciding
to other improved stoves, lack of awareness of with the December holiday. The transition phase
efficient cooking electric appliances and the in January 2023 coincided with the resumption of
diverse electric appliance options, limited distribu- work activities in the same month.
tion points, and the cost of electricity.
Details of the E-cook Stoves provided in
In response to these barriers, innovative inter- the transition phase
ventions have been rolled out in various markets,
such as Kenya, that include the cost reduction In the transition phase, households were provided
of devices, credit facilities, utilities-led financing, with 2-3 appliances comprising an EPC, hotplate,
carbon finance, and results-based financing (RBF), and an optional air-fryer. These appliances were
which have led to the emergence of the use of the most available based on their local availability,
multiple electric appliances at the household level. and their socio-cultural compatibility was assessed
Suppose these interventions in Kenya may result per the cooking diaries protocols using methods
in a 100% uptake of electric cooking in the future. outlined therein. The choice of providing the
In that case, there is a need to understand if the optional air fryer to a household was made consid-
current e-cook technologies and cooking practices ering both data about food cooked in the baseline
can support a 100% e-cook scenario. This study will and assessing the load-handling capability of the
further provide researchers visibility of barriers and wiring in the household.
future opportunities at scale and provide evidence
in support of new interventions. Overview of changes noted in meals and
dishes cooked.
This study is the first assessment of households
cooking exclusively using electric cooking Data captured across both phases indicates
appliances in Kenya through the lens of 10 urban that the main meals cooked among the selected
households in Nairobi, with low, and middle-in- households are supper, lunch and breakfast, repre-
come households equally represented. This study senting 96.1% of all dishes cooked, with baby food
is part of a series of studies on the implication of and snacks comprising 3.9% of the meals cooked.
households cooking entirely with electricity by the Of all meals, supper had the most dishes cooked
Modern Energy Cooking Services (MECS). This in both study phases, cementing supper’s constant
study utilizes a modified form of the cooking diaries importance across both holidays and workdays.
framework version 3.0 to support the assessment Transitioning to electric cooking using multiple
of multiple e-cook appliances in the household, appliances did reveal an increase in the number
compared to previous studies that issued a single of dishes cooked per meal. This change may be
intervention appliance per household. The cooking attributed to a combination of factors such as ease
diaries framework is used to investigate the degree of EPC use and the positive effect of the electricity
of compatibility of electric cooking devices with subsidy encouraging greater use of electricity but
local menus. It presents this objectively through as- conversely leading to an unintended increase in
sessments of foods cooked, energy consumption, the number of dishes cooked per meal.
and cost (relative to traditional fuels).
ix
The cooking diaries consider three dish states, foods to transition to electric cooking. Data revealed
namely, fresh dishes, partially cooked dishes, that the proportion of ugali cooked on LPG increased
and reheated dishes. The consultant found that from 14.8% to 19.7% in the baseline and transition.
households mainly cook fresh dishes, with the Charcoal data usage shows the most popular
most significant number of total dishes at 64.1% usage of charcoal in the baseline was cooking meat
in the baseline and 54.5% in the transition phase. and chapati. However, during the transition phase,
Reheated dishes had the second highest signif- charcoal was only utilized to prepare cereals. The
icance in the study, accounting for 30.0% and continued use of charcoal, even with the availabili-
32.8% in the baseline and transition phases, re- ty of EPCs in the transition, may be attributed to the
spectively. Partially cooked dishes represented batch cooking of large amounts of grains obtained
the least occurring dish state. The consultant did by households from relatives in rural areas during
note a change in dish state in breakfast and lunch the December holidays.
meals, with the baseline data showing fresh dishes
were most popular but was surpassed by reheated This discovery shows that current electric cooking
dishes in the transition phase. This change is not technologies, such as EPCs, are still unsuited for
attributed to the introduction of electric cooking batch cooking activities. As a result, electric cooking
but may be attributed to the resumption of work appliances, as is, may not wholly displace the use
activities in January, which reduces the time of traditional fuels such as charcoal in the event of
available for cooking. such a batch cooking scenario.
Overview of changes noted in fuels used Overview of changes noted in cooking

time, energy, and fuel costs.
Across both research phases, households utilised
an assortment of fuels in varying proportions. In the transition, households were provided with
Baseline fuel usage showed LPG as the most used locally available EPCs, and hotplates, with air fryers,
fuel, with a usage of 76.0%, followed by electricity only provided to households that could sustain the
at 21.9% and charcoal at 2.1%. Most of this electrici- additional electrical load.
ty usage in the baseline was mainly from reheating
utilizing microwaves and not actual cooking using Data on cooking time at the dish level showed that
other electric cooking devices. In the transition the lengthy time taken to cook a dish on a hotplate
phase, electricity was the most used fuel with a combined with the pre-programmed cooking time
usage ratio of 86.3%, while LPG use dropped to on the EPC and air fryer might not yield time savings
13.5%, and finally, charcoal at 0.2% of all dishes while cooking fresh dishes. However, electric
cooked. cooking does yield time savings for reheated
dishes due to the faster reheat time while utilizing
It is evident that despite the motivation to use the microwave. Despite the cook-time concerns,
electricity, households still cooked using LPG and the automated cooking on an EPC does deliver a
charcoal, necessitating a further analysis of the more straightforward cooking process as the EPC
foods cooked using these fuels in both phases of can be left unattended.
the study.
At the meal level, the consultant assessed if the
LPG usage data in the baseline showed that ugali non-improvement in dish cooking time would have
was the most popular meal, followed by rice, beans, the same effect on meal cooking times. Meal data
an array of meaty dishes, leafy green vegetables, collected shows that breakfast and lunch meals
and tubers. Ugali was the most difficult dish of these
In the transition phase, electricity was

the most used fuel with a usage ratio
of 86.3%, while LPG use dropped to
13.5%, and finally, charcoal at 0.2%
of all dishes cooked.
x
witnessed reductions in cooking time of 47.7% Perceptions and barriers to cooking

and 40%, respectively. Supper data shows a 20.1% entirely with electricity.
increase in cooking time compared to the baseline.
This reduction in the time to cook breakfast and The study assessed the barriers to cooking with
lunch meals may be attributed to microwave utili- electricity, shown in Table 1 and Table 2. In addition
zation in reheating dishes in the transition phase. to the barriers, perceptions of the experience of
The use of microwaves in heating can be further cooking with electricity were noted as follows:
attributed to the resumption of work activities
in January, which reduces the time available for Heat regulation capability– The EPC and air
cooking. Supper cooking times show a 20.1% fryer received 67% and 75% approval, respec-
increase due to multiple dishes cooked in supper tively, and the hotplate received 80% approval
meals, with most supper meals comprising two across those households that received each
dishes. However, the increase is not as profound appliance type
due to the complementary use of LPG (especially
those with multiple burners), charcoal, and electric Satisfaction with the cooking pots – The EPC
appliances in the transition phase. This evidence received a 78% approval rating due to the
was further compounded by the study’s constraint, sizeable 8L volume of the pressure cooker. The
where each household only received a single EPC, 3.5 L air fryer obtained a 40% approval rating
hotplate and possibly an additional air fryer. Due to due to the 3.5 L volume. However, a 5.5L of the
the low thermal coupling between the hotplate and same brand exists. The inability of the hotplate
the cooking utensils used and safety concerns with to accommodate the typical sizes of cooking
its use, electricity outages, and user perceptions of pots owned by the household resulted in the
electric cooking, participants could not exclusively hotplate receiving a 40% disapproval rating.
use electric appliances.
Speed of cooking – The highest approval
A straight-line model was developed to determine ratings were conferred to both the EPC and air
the effect of transitioning from baseline fuels to fryer obtaining 78 and 75% approval ratings, in
electric cooking. The model describes an electric contrast to the hotplate, which received 89%
transition’s per capita energy and per-capita cost disapproval across all households.
dynamics in the study group. A complete lack of use
of electric cooking within the study group would Safety – Participants rated the EPC and the air
result in a per capita cost of 0.13 US$ per dish from fryer as having a high safety index. However, the
using LPG and charcoal, while 100% e-cook results hotplate received unfavourable reviews across
in a financial cost per capita of 0.02 US$ per dish. all households due to the exposed heating
A household would have an average cost saving element. One specific household did not allow
per dish of 0.11 US$ per capita per dish cooked. In children to cook with a hotplate without adult
addition to energy costs, a lack of use of electric supervision.
cooking appliances within the study group would
result in a per capita energy use of 3.68MJ per dish Ease of use and training – As part of the EPC
from using LPG and charcoal, while 100% e-cook accessories, an e-cookbook was provided to all
results in a per capita energy requirement of 0.41MJ households. No guidebook was supplied with
per dish. A household would save an av-erage of the generic hotplate, while the households that
3.27 MJ per capita per dish cooked. With an energy received the air fryer received only a standard
cost of 0.11kWh (0.41MJ) per person per dish, while product manual. All households mentioned
cooking entirely with electricity, a household with that training on electric appliance use was
2.9 household members, such as those in Nairobi not strictly required. A positive highlight was
(KNBS, 2019), that cooks one to two dishes per meal the appreciation of the e-cookbook, which
daily across breakfast, lunch, and supper would allowed the households to start using the EPC
consume between 29.82 and 59.65 kWh of elec- by following the e-cookbook guidelines on how
tricity monthly. to cook select local dishes using the EPC.
xi
The e-cookbook permitted the transition to Further suggestions to manufacturers –Most

proceed without a steep learning curve. of the comments centred around the cooking
pots with the idea that a single EPC pot was not
Opinions comparing electricity vs LPG use – enough and households required multiple EPC
Participants singled out LPG’s reliability as its pots to at least perform sequential cooking if
most outstanding advantage over electricity. parallel cooking was not economically feasible
However, there were several disadvantages for the manufacturer. The lack of multiple pots is
associated with the use of LPG. These included compounded by other results that show supper
the constant need to purchase matchboxes meals comprised an average of two dishes,
that usually get misplaced and the inability requiring more than the single pot provided
to determine how much LPG remains while by EPCs. The consultant further noted that the
cooking. In addition, participants mentioned that households use several cooking devices simul-
cooking using LPG was faster than using the taneously during a cooking event.
electric hotplate. However, electric appliances’
ability to cook multiple dishes simultaneously Therefore, the consultant determined via
as in a multiple-burner LPG stove is a desired the exit survey that a household requires 3.4
feature. However, current technologies do electric cooking hobs to allow the parallel use of
not permit this, resulting in electric appliances electric appliances. With an EPC only providing
only being able to cook a single dish at a time. a single cooking hob, there is an opportunity
Despite this limitation, providing extra pots to introduce electric cooking appliances with
on the EPC would somewhat overcome this multiple heating elements to compete with
hurdle. All households except one agreed that multiple-burner LPG stoves. Additionally, it was
it was more affordable to cook with electricity suggested that hotplates should be designed
compared with LPG to accommodate non-flat-bottoms and larger
pots.
Table 1: Summary of barriers

# Category Specific
1 Supply of electricity Unreliable supply of electricity (frequent black-outs)
Shared electricity meters in some rental houses
2 Electric-cooking Slow cooking time for the hotplate case
appliance limitations
Inability to cater for large quantities of food
Inability to cook multiple dishes simultaneously on one appliance
Requirement for a specific cooking pot for the EPCs
Inability to cater for round/large pots for the case for a hotplate
Safety concerns for the hotplate
Complex procedures for a small dish
Incompatibility with some local dishes, e.g. chapatis for EPCs
3 User-Limitations Lack of awareness of the financial benefits of cooking entirely with
electricity
Beliefs that certain meals are better cooked on a specific cooking solution,
e.g. ugali on LPG stoves
User preference on a specific cooking solution, e.g. LPG
xii
Table 2: Frequency of reasons for not cooking with electricity

# Reason Frequency (N)
1 Hotplate too slow 20

2 Blackout 17
3 Too much to cook 10
4 No prepaid energy balance 7
5 Too little to be cooked 6
6 LPG cooks better 5
7 Cook forgot 4
8 Cook was in a hurry 3
9 Inability to use electric appliances to cook 2
10 The air fryer takes smaller portions, hence moved to the gas 1
11 Lack of adult supervision 1
Conclusions and recommendations

1. Transitioning completely from baseline fuels (A towards modelling a special e-cooking tariff for
combination of LPG, charcoal, and electricity) to the 2024-2025 electricity tariff review period,
100% electric cooking in the transition scenario and beyond.
with EPCs, hotplates and air fryers will reduce its
energy use from 3.68 MJ to 0.41MJ/person per 2. A cooking book tailored to local meals by a man-
dish. Such a transition would also reduce the ufacturer/distributor is preferred to a generalised
per capital cost per dish from 0.13 US$ to 0.02$ manual on operating electric appliances.
per person per dish. Translating these results
to urban households in Nairobi shows that a 3. With the requirement of 3 - 4 parallel-run (The
typical household with 2.9 house members, as ability to cook more than one dish at a time),
per the 2019 census, that cooks one dish per the availability of enough electric appliances
meal consistently across three meals a day ex- to meet cooking requirements was not
clusively using electric appliances would have possible in the study. As a result, households
a monthly electricity usage of 29.82 kWh on used standard fuels in parallel with electric
cooking activities. This usage is barely within appliances. Additionally, the study found that
the subsidized lifeline tariff band (0-30 kWh at electric appliances, including EPCs, are not
12.22 KES/kWh) in the 2023-2024 KPLC tariffs. yet suitable for performing large-batch cereal
As part of its tariff review notes for this period, pre-cook activities, resulting in the utilization of
the energy regulator, EPRA, mentioned the lack charcoal in such cases. There is, therefore, a
of sufficient available data to allow the modelling need to extend the scope of this study to inves-
of a specific tariff for e-cooking. We believe that tigate these findings further.
the models presented herein build a strong case
xiii
About
1 billion
in Sub-Saharan
Africa (SSA) rely
on biomass fuel,
kerosene, and
coal for cooking
©Shutterstock
xiv
01
Introduction
1.1 Background
About 2.4 billion people globally and about 1 The low uptake of cooking with electricity has
billion in Sub-Saharan Africa (SSA) rely on biomass been attributed to the high up-front cost of electric
fuel, kerosene, and coal for cooking1. It is further stoves compared to other improved stoves, lack of
estimated that 2 billion people will still have no awareness of efficient cooking electric appliances
access to clean cooking solutions by 2030 if there and the diverse electric appliance options. Limited
is no global political will and concerted efforts from distribution points and the high cost of electricity
the different stakeholders2. Cooking using tradi- in Kenya have also been cited as factors limiting
tional forms of biomass, especially fuelwood and e-cooking uptake6.
charcoal, is now the leading source of greenhouse
gases in Sub-Saharan Africa3. This practice is also Adoption of these solutions remains low in Kenya
associated with widespread negative environmen- and Sub-Saharan Africa due to a complex mix of
tal and health impacts. these factors that can be conceptualized as de-
mand-side and supply-side barriers.
Although Kenya’s electrification rate as of 2018 was
75% (53.5% have access through the national grid, The triggers, enablers, and causes of transitions
while 21.5% use off-grid solutions)4, the use of elec- in cooking fuels at the household level have been
tricity for cooking is still low. According to a national studied extensively over the last three decades.
survey by the Ministry of Energy in 2019, approxi- Many studies demonstrate a correlation between
mately 3% of the households owned a mixed LPG- a household’s socioeconomic status, a proxy
electricity stove, electric coil and microwave. Table indicator of disposal income, and its cooking
3 below provides information on electric cooking options7,8,9.
solutions in Kenya, identified in 20195.
Table 3: Use rates of electric appliances under the Kenya Cooking Sector Study (MoE,2019). Source: EED
Aggregate category Specific category % of households using electric appliances
Urban Rural National
Electric cooking appliances Mixed LPG-Electricity Stove 4.9 0.3 1.8
(Approx. 3% use rate)
Microwave 2.2 0.2 0.8
Electric coil stove 0.8 0.0 0.3
Electric induction stove 0.0 0.0 0.0
1. World Bank. (2021). Tracking SDG7 the Energy Progress Report 2021.Retrieved from https://iea.blob.core.windows.net/assets/8b276fc2-
c1ae-4a54-9681-eea1eb143d7f/TrackingSDG7TheEnergyProgressReport2022.pdf
2. Ibid
3. Bensch, G., Jeuland, M., & Peters, J. (2021). Efficient biomass cooking in Africa for climate change mitigation and development. One Earth,
4(6), 879–890. https://doi.org/10.1016/j.oneear.2021.05.015
4. ESMAP. (2019). Kenya Beyond Connections Energy Access Diagnostic Report Based on the Multi-Tier Framework. https://energydata.info/
dataset/kenya-multi-tier-framework-mtf-survey/resource/b27a89d9-118c-4139-b1ad-70d5abd0b508?view_id=1591cd89-790f-4b2d-9f5f-
66022b3
5. Ministry of Energy. (2019). Assessment of the Supply and Demand of Cooking Solutions at Household Level.
6. Ministry of Energy. (2019). Assessment of the Supply and Demand of Cooking Solutions at Household Level.
7. Hosier, R. H., & Dowd, J. (1987). Household fuel choice in Zimbabwe: An empirical test of the energy ladder hypothesis. Resources and
Energy, 9(4), 347–361. https://doi.org/10.1016/0165-0572(87)90003-X
8. Masera, O. R., Saatkamp, B. D., & Kammen, D. M. (2000). From Linear Fuel Switching to Multiple Cooking Strategies: A Critique and Alter-
native to the Energy Ladder Model. World Development, 28(12), 2083–2103. https://doi.org/10.1016/S0305-750X(00)00076-0
9. van der Kroon, B., Brouwer, R., & van Beukering, P. J. H. (2013). The energy ladder: Theoretical myth or empirical truth? Results from a
meta-analysis. Renewable and Sustainable Energy Reviews, 20, 504–513. https://doi.org/10.1016/j.rser.2012.11.045
1
Approximately
3%
of the households
owned a mixed
LPG- electricity
stove, electric coil
and microwave.
©Shutterstock
2
Hosier and Dowd (1987) promoted the energy adoption include the high upfront cost, limited
ladder theory, which notes that as a household awareness, socio-cultural practices, limited access
moves up the socioeconomic ladder, traditional to suitable options, high cost of electricity, and
fuels are replaced by transitional fuels. low-quality electricity supply. That households
will automatically transition completely to electric
Eventually, these are replaced by advanced fuels. cooking if these barriers are removed remains an
Masera et al. (2000), however, argue that although assumption that needs to be tested and further
households transition from one fuel type to the studied. Providing electric solutions and affordable
other as their progress up the socioeconomic electricity to households could result in transitions
status, they do not entirely abandon primary and including: i) complete and immediate, ii) complete
transitional fuels but use this together with modern but gradual, iii) partial and immediate, iv) partial but
fuels in a stack. This observation was popular- gradual, or v) no transition.
ized as the energy matrix theory. With the need to
accelerate access to clean cooking solutions under Household type (e.g., standalone versus multiplex
SDG-7 and Kenya’s ambitious goal of achieving units), setting (e.g., urban versus rural), income
universal access to clean energy cooking services level, characteristic (e.g., family versus individual),
by 2028, understanding the drivers of choice in cooking requirements, availability of alterna-
cooking solutions at the household level is now tive cooking solutions, and other attributes will
more urgent. inevitably influence the type of transition.
There is a rapid evolution in electric cooking tech- The Cooking Diaries methodology developed by
nologies, including new designs of electric pressure MECS generates dish-level qualitative and quan-
stoves, air fryers, rice cookers and electromag- titative data on the cooking practices within a
netic induction stoves in the Kenyan market. This household, including the amount of energy, cost,
evolution is complemented by increased access and time used. The methodology is, therefore, also
to electricity and innovative business models, such suitable for investigating the compatibility of electric
as pay-as-you-go solutions that resolve the afford- cooking devices with various household contexts.
ability challenge.
MECS has commissioned studies investigating the
implications of households cooking entirely with
These factors combine to create an opportunity
electricity in multiple countries, including Kenya,
that supports the transition from traditional forms
using a portfolio of electric cooking device options
of cooking to cleaner and, in some cases, more
within a household. The research will be guided
affordable electric cooking options. Simulating
by the Cooking Diary protocol version 3.0 with the
and documenting a complete shift to electric
primary objective of understanding the implications
cooking is an essential first step in understand-
of cooking entirely with electricity at the household
ing the challenges and opportunities available
level and how these options can be used to meet
to households as they adopt cleaner cooking
all household cooking needs.
solutions. It is conceivable that even with access
to electric cooking solutions and affordable elec-
The specific objectives of this study are to determine
tricity, some households may not be only willing or
the following:
able to transition to electric cooking. Others may
transition only partially. Studying the implications i. Energy implications of a household cooking
of a complete transition will address some of these entirely with electricity.
knowledge gaps.
ii. The benefits of transitioning to cooking entirely
with electricity include cost and time savings.
1.2 Purpose of the study
iii. Barriers to transitioning to cooking entirely with
electricity, such as cultural contexts.
Displacing traditional forms of cooking with electric
solutions has multiple health, climate, economic, iv. User experience and user preferences while
and environmental benefits. Core barriers to cooking entirely with electricity.
3
02
Approach and
Methodology
This study investigated the implications of households cooking entirely with electric-
ity using a portfolio of electric cooking devices. The research was guided primarily
by the Cooking Diary Protocol version 3.0. It employed the approaches comprising:
i. assessment of the electric cooking devices,

ii. data collection, and
iii. synthesis and reporting.
A summary of activities and outputs is in Figure 1 below.
Figure 1: Summary of Approach
1 2 3
Assessment of electric Data collection Synthesis and reporting
cooking devices
• Rapid mapping of electric • Obtained a research • Data cleaning
cooking devices available permit • Data Analysis
in Kenya • Selected participants • Develop PPT with
• Determination of everyday • Selected enumerators summary of analysis
dishes • Conducted entry survey • Develop a study report
• Asses electric devices’ • Implemented the
compatibility with standard baselines survey
cooking processes • Conducted exit survey
• Electric cooking devices • Carried out Focus group
availability report discussion
2.1 Assessment of the available electric cooking devices

2.1.1 Electric cooking devices available in the Kenyan market
An electric cooking appliance availability assessment 248). The online search process involved an
was undertaken to determine available electric in-depth search of online retailers that deliver
appliances that would assist the households in tran- directly to Nairobi County customers. The
sitioning to cooking entirely with electricity. The consultant searched for the websites of online
consultant used the following processes to survey retailers selling electrical cooking appliances.
the availability of electrical cooking appliances: To perform a similar approach as the super-
markets, we excluded appliance models that
An online search of prominent appliance were out of stock. The online search focused on
vendors with an online B2C presence – The electric cooking appliances by capturing the ap-
consultant performed a search for available pliance’s brand, model, power rating, price, and
electric cooking appliances to determine image.The consultant carried out this activity
appliance costs and bound the appliances between the 10th -15th of November 2022.
sought to an upper limit of 30,000 KES (US$
4
Visits to supermarket chains at specific 2.1.2 Determination of common dishes

locations– The consultant followed the same A variety of dishes are prepared across households
approach with online retailers. This process in Kenya. The different dishes or meals need
involved an in-depth search of retailers that specific cooking requirements and processes. For
deliver directly to the customer in Nairobi example, ugali, a staple food in Kenya, requires
County. This process involved visiting retailers boiling water in a cooking pot and adding maize
and the information captured on an excel sheet, flour, rigorous stirring until it is evenly cooked with
such as the electric cooking appliance’s brand, a soft to medium-hard texture. The cooking process
price, size, and power rating, where available. for ugali differs from other foods, such as githeri,
The consultant performed the visits between which involves boiling beans and maize separately
the 10th -15th of November 2022. to prepare and soften them. After this, they are both
shallow fried with other ingredients for flavour. The
consultant grouped the most common among the
10 (ten) selected participants and presented this
Figure 2: A bar graph of the most common dishes cooked by data in Figure 2.
the ten selected households.
10 10 10 10
9 9 9 9 9 9
8
Number of Households
7 7
6 6 6
5 5 5 5
3 3 3
0
Irish potatoes
Dried cow peas leaves

Rice
Sukuma wiki
Cabbage
Sweet potatoes
Chapati
Water for tea
Spaghetti
Beans
Porridge
Dried vegetables
Dried veg with G.nuts

Ugali
Beef
Egg
Free range chicken
Cake
Githeri
Cow peas
Dried fish
Fresh fish
Cassava
Chicken broiler
Matoke
dish/ meal
5
2.1.3 Compatibility of available e-cook appliances with cooking processes

Different foods have unique techniques of prepara- that regularly fry and roast food. Its selection and
tion. These processes determine the most optimal use in the study were based on the lack of an al-
cooking appliance used to prepare the food. Electric ternative for these types of meals. The electric hot
pressure cookers, hot plate cookers, air fryers plate cookers, though inefficient, can cook various
and tabletop gas cookers with electric hotplates meals. While a tabletop gas cooker with a hotplate
emerged as the optimal combination. An electric can be considered if the electric function is in use,
pressure cooker is suitable for boiling heavy foods such an option was still left out. This omission was
such as githeri, beans only, beef, chicken, and done to avoid the likely confusion in the reporting,
cowpeas. mainly where the LPG functionality was used. As
a result, the most optimal combination of electric
Additionally, it is ideal for cooking staple foods appliances selected included the electric pressure
that require boiling, such as ugali, Irish potato, rice, cooker, electric hot plate cooker and air fryer. The
sweet potato, egg, matoke, spaghetti, and porridge. specific brands of e-cook appliances provided are
The air fryer’s convective cooking process is a shown in Table 4 below.
suitable alternative to deep-frying for households
Table 4: E-cook appliances chosen to be provided. Source: EED
No. Type of electric stove Model Cost

1 Electric pressure cooker (8L) ECOA KES 9500 ($ 78)
2 Electric hot plate (single coiled burner) Generic KES 1,000 ($ 8)
3 Air fryer (3.5 L) VSYA35BBK KES 8,995 ($ 74)
6
2.2 Data collection

2.2.1 Pre-data collection activities
i. Application for a license to undertake Household income level categorisation: To
research have an equitable study across income groups,
the consultant selected five (5) low-income
The consultant applied for a research license from and five (5) middle-income households. The
the national research regulator, NACOSTI, per Kenya National Bureau of Statistics classifies
the Science, Technology, and Innovation Act of a household as a low-income household if
2013, which requires a permit before undertaking its monthly spend is KES 46,355 (US$ 382)
research activities. The application form captured or less per month, while a middle-income
all aspects related to the project’s objectives, household is one whose monthly spend is
including but not limited to the research structures within the range of KES 46,356 ((US$ 382) and
involved, the profile of the principal researcher, and KES 184,394 (US$1,518) per month10.
the client. A week after the application filing, the
regulator issued the permit allowing the consultant Baseline stove(s): According to the Kenya
to kick-start the recruitment of participants. household cooking sector study, the top
three popular cooking solutions in urban
areas include LPG solutions (54.4 %), charcoal
ii. Selection of households to participate in
stoves (47%), and kerosene stoves (27%)11. The
the survey
consultant sought households that use at least
The consultant created an evaluation criterion to one of these stoves.
assess households’ alignment with the project’s
primary objectives and to generate high-quality Household size: According to the 2019 census,
data. These criteria aimed to obtain an initial set of the average household size in Nairobi is 2.9 (3
households that would undergo further verification members). To test the limits of electric cooking,
to assess their suitability. the consultant sought larger households as
they would cook most meals due to home
The evaluation criteria for the initial set of cooking being more economical than
households included the following: restaurant foods.
Access to electricity and quality of electric

wiring in the kitchen: Additionally, the
consultant required that the households be
connected to electricity, have proper wiring to
accommodate the electric cooking load, and
have sockets in the kitchen.
Motivation: The consultant sought motivated

households that would be interested in the
study and that would cooperate with the enu-
merators.
Location: Based on the consultant’s require-

ments and for ease of administration, the
selected households were near the con-
sultant’s work premises. Kawangware was
designated as the study area in Dagoretti
North sub-county in Nairobi City County. The
study area is illustrated on the map in Figure
3.
©Shutterstock
10. Kenya National Bureau of Statistics (2022). Economic Survey 2022. https://www.knbs.or.ke/wp-content/uploads/2022/05/2022- Econom-
ic-Survey1.pdf
11. Ministry of Energy (2019). Kenya Household Cooking Sector Study: Assessment of the Supply and Demand of Cooking Solutions at the
Household Level
7
Figure 3: Contextual map of the study area, Kawangware, in Dagoretti North Constituency in Nairobi City County
8
Source: EED Advisory
Fourteen (14) households were selected using the consultant would not be able to attribute the
snowball sampling. Further, the consultant analyzed change in weight in the LPG canister to a specific
quantitative and spatial data from these households dish. The cooking diaries protocol version 3.0
using MS Excel (Microsoft, Redmond, Washington, advises that all households use large LPG cylinders,
USA) and QGIS (QGIS Development Team, Global). such as the 13-kg LPG cylinder, to simplify LPG mea-
The quantitative data included a description of surement. Therefore, households using a mixed
the dishes cooked, household income level, LPG-electricity stove with multiple burners as
and cooking patterns. On the other hand, spatial their primary energy source were given two 6-kg
data captured the location of the households in complete LPG cylinders. Some households with a
Kawangware. Of the fourteen (14) households, ten 6-kg complete LPG cylinder as their main energy
(10) met the requirements for inclusion in the study. source were given an additional 6 kg LPG cylinder
depending on their parallel cook needs. The 6-kg
LPG cylinders were bought from reputable vendors
iii.Determination of fuel measurement
within the study area, and the enumerators dis-
strategies
tributed them to the households before the com-
a.LPG, kerosene, and firewood measurement mencement of the baseline survey.
An initial assessment of the cooking devices Firewood use measurement via weight involves
used within the households revealed that some collecting the firewood into a bucket and measuring
households used either a 6-kg complete LPG the firewood before and after use utilizing the
cylinder, mixed LPG-electricity stove with multiple mechanical support tool in Figure 4. However, a
burners (connected to 13-kg LPG cylinder), or both smaller bucket was utilized for charcoal measure-
for cooking. ments, while kerosene, usually stored in bottles,
would be measured directly.
Measuring the amount of fuel used for cooking
different dishes was critical to the study. The As kerosene may also be stored within the fuel
consultant discovered that measuring a 13-kg LPG compartment of the stove, kerosene measure-
cylinder after every meal proved too difficult for the ment procedures also involve measurement of the
participants. Additionally, using stoves with multiple stove before and after cooking. At the end of the
burners meant that parallel cooking on multiple study, the consultant noted no use of kerosene was
burners would complicate the data collection as recorded within the study across all households.
Figure 4: Mechanical support for LPG, kerosene, and firewood measurements. Source:EED
9
b. Bio-ethanol measurements c. Electric cooking appliance measurements
The consultant noted the use of ethanol in Koko The consultant utilized the Sonoff POW Elite in
cookers (Koko Networks, Nairobi, Kenya). Due to Figure 5 and Figure 7 due to its combined energy
the bio-ethanol’s evaporative nature and the Koko meter and data logger features. A downside to its
cooker’s shape, it was determined that the most use discovered after shipping is that the LCD screen
appropriate means of measurement of ethanol resolution was 1 kWh and not 0.1 kWh, as advertised.
used on Koko cookers was through the Koko app. However, the consultant found that the data trans-
The logic behind this was that Koko had invested mitted over Wi-Fi was accurate to 0.01 kWh as
in accurate weights and measures of their fuels. shown in figure 5. As a result, an effort to use this
As a result, the consultant determined that their data began by implementing the data architecture
systems could determine the amount of ethanol to shown in Figure 6. Wi-Fi outers were purchased
disburse to a client. This amount and the account for all households with Wi-Fi SSIDs and passwords
statement were useful in computing energy costs. hardcoded per appliance monitor in advance to
However, a temperature logger was installed to allow a seamless installation experience.
determine the proper fuel use per dish, and its data
was retrieved at the end of the study. The tempera- Figure 5: The Sonoff app. Source:EED
ture data would allow the time division of the fuel
used between purchases. However, this does not
account for varying fuel flow adjusted using the
stove’s fuel regulator.
Despite the efforts toward measuring ethanol use,

the consultant noted that the households did not
fully utilize these cooking devices once the shift
was made to the 6-kg cylinder. This change was
attributed to the greater convenience of LPG once
the cost factor was removed.
Figure 6: Data flow of minute-by-minute electrical measurement data. Source:EED
Consultant’s software on Linode

1TB S3 storage
Ewelink (Manufacturer)
Data Data
Data
Data
Sonoff Elite on Router Enumerator access of

Electrical cooking
consultant’s tools on
appliance
Google Colab
Household boundary
10
Each router contained a SIM that a third-party Figure 7: Sonoff POW Elite monitoring an EPC. Source:EED
SIM management company managed, allowing
the consultant to ensure 100% connectivity. The
consultant wrote special software to run on a
leased server on Linode (Linode LLC, Philadel-
phia, USA), a cloud service provider. The software
made an API call to Sonoff POW’s manufacturer
software system named eWelink (Shenzhen Coolkit
Technology, Shenzhen, China), which allowed the
collection of both voltage and energy measure-
ments and other metrics every minute across the
four-week study. The retrieved data was then
stored in an online S3 storage as individual files,
whose timestamp uniquely identified the mea-
surement. The consultant further utilised Docker
(Docker Inc, San Francisco, USA) to ensure that the
software tools developed would restart themselves
automatically if unhandled software errors were en-
countered. This strategy proved very successful,
with only one data outage on the consultant’s
systems within the entire study period.
Further, the consultant developed the inspection

tools, shown in Figure 8, using Google Colab
(Google, California, USA)to allow the enumera-
tors to use this collected data and compare it with
the field data on appliance usage in the form of a
cooking diary, filled in by the households.
Figure 8: EPC energy usage within the transition phase on Google Colab
25
Cumulative kwh consumed (units)
20
15
0
Jan 5 Jan 7 Jan 9 Jan 11 Jan 13
2023
Date and time of day (Nairobi time)
11
The enumerator, therefore, would access the tool 3. Introduction of the cooking diary forms: The
daily and verify the measurements before filling consultant trained the enumerators on the
in the initial and final energy in kWh into the Kobo cooking diary form. This form was filled out
cooking diary forms. This data entry strategy was by the household at each dish-cooking event.
pivotal in recognizing discrepancies in household The focus on the dish is performed because
data. Where the enumerator discovers a discrep- fuel consumption is directly proportional to the
ancy, the consultant would send a note informing heat generated from its combustion. This heat
the household of the discrepancy and seek to is typically applied to a cooking pot containing a
understand further the root cause. single dish. As a result, fuel use can be directly
mapped to dishes prepared.
iv. Recruitment and training of enumerators
4. Training on the registration survey (survey
For adequate supervision, the consultant required informing the selection of enumerators): This
2 (two) female enumerators that would each training involved understanding the research
administer one of the two clusters of households. questions within the survey questionnaire.
This administrative structure simplified the inter-
action between the participants and the enumer- 5. Training on the data collection phases: The
ators, simplified data collection and allowed quick consultant trained the enumerators on the
relaying of challenges raised by households to the baseline and transition data collection phases,
consultant. Additionally, the consultant ensured that per the study objectives.
the participants were comfortable with the enumer-
ators’ daily visits to collect the forms and observe 6. Training on weighing the fuels and recording
how the participants were recording data. the energy meter readings: The consultant had
sample weighing scales and electric meters in
After recruitment, the consultant engaged the enu- the training venue and trained the enumerators
merators in training that entailed several training on how to use them.
modules:
7. Training on data collection etiquette: The
1. Introduction to the study: The introduction consultant trained the enumerators on best
highlighted the background, purpose, and practices when dealing with the participants
expected outcomes. within their households due to the daily visits to
the households.
2. Introduction of the survey tools: The enumer-
ators were already familiar with the Kobo collect 8. Provision of survey tools: The consultant
survey tool as the consultant had engaged them provided data collection tablets and trained the
on past data collection assignments. As a result, enumerators to use them. A mock test was finally
they understood how to use the tools. performed after the training.
©Shutterstock
12
2.3 Implementation of the surveys

Four surveys were implemented during the study period. A summary of each survey is in Figure 9,
with further information discussed below on each survey type.
Figure 9: The main data collection steps of the study
Entry survey Baseline Survey Transition Survey Exit Survey

1 day 14 days 14 days 1 day
Main activities Main activities Main activities Main activities

• Administered entry survey • Participants filled cooking • Participants filled cooking • Administered exit survey
forms diaries forms. diaries forms. forms
• Participants registered for • Monitored participants • Monitored participants • Assessed participant’s
the study. cooking habit with cooking habit with electric perception toward
preferred fuel & cooking cooking appliances. cooking with electricity
• Ensured data quality
appliances. - FGD
by checking possible • Provided subsidy for
incorrect entries • Measuring different fuels participants (electricity • Ensured data quality
token)
• Ensured data quality
by checking possible • Ensured data quality
incorrect entries by checking possible
incorrect entries
2.3.1 Entry survey

The baseline began with a stacked week
The entry survey (ANNEX 2) sought to understand (households cooking with multiple cooking
the participant’s fuel usage patterns, cooking technology or fuel). During this stack week, the
devices and habits. It also aimed at collecting user consultant gave the participants the liberty of
metadata that would allow efficient processing of cooking with different fuels during meal prepara-
the data collected in the analysis phase. The entry tion. During the second unstacked week, the partic-
survey was a standard document provided by the ipants were encouraged to cook with their primary
MECS team and was utilized as was provided, with baseline fuel to collect as much meal and dish
minor logical flow changes. data as possible. The stacked week data would
allow cooking comparisons entirely with electrici-
The entry survey was administered on 05 ty (transition phase) to primary baseline fuel. As a
December 2022. result, the final unstacked week requires that the
household cooks all dishes and meals using the
primary fuel, where possible.
2.3.2 Baseline phase
Following the entry survey, households were
to cook using everyday cooking solutions. The 2.3.3 Transition phase
baseline commenced on 06 December 2022 and Toward the end of the baseline phase, participants
was carried out using the cooking diary (ANNEX were invited to an electric cooking training on 22
1) form coded and integrated into Kobo Toolbox. December 2022 in preparation for the transition
The enumerators ensured that all households phase in January 2022. During the training, the in-
understood the cooking diary form, procedures for structors demonstrated how hotplates and EPCs
measuring the fuels right before and after cooking are used in cooking. This training is shown in Figure
use, and time recording before and immediately 10. Further, videos were recorded during training
after cooking, among other pertinent issues. Par- and uploaded to the WhatsApp groups of the par-
ticipants and enumerators were allowed to interact ticipants for reference.
with cooking diary forms and other equipment
issued to them for one week before the baseline On resumption of the transition phase of the
started. All the enumerators and participants were study in January 2023, the consultant provided
monitored during this period to ensure high-qual- the households with a 5-day buffer to familiarise
ity data. themselves with the e-cook appliances.
13
After five (5) days, the transition phase began questions posed. As such, an analysis guide is used
officially for two weeks. In this phase, the consultant to achieve the result as it specifies how to dissect
expected households to cook all their meals with the data.
electricity (100%), where possible.
As a result, not all analyses were used from the
cooking diaries analysis guide but only those
2.3.4 Exit Survey related to the study objectives.
At the end of the transition phase on 22nd January,
the consultant conducted an exit survey (ANNEX The dissection of the data was carried out using
3) across the households to get their perspec- Microsoft Excel and Pandas tools. However, only
tives about cooking exclusively with electricity. In Microsoft Excel was used during the data prepara-
addition, FGDs were conducted to reinforce the tion phases. Pandas were used in the later stages
quantitative data obtained during fieldwork and due to their versatility in creating functions that
fill gaps in case there are missing data points from could be reused even if the raw data changes.
quantitative data. The data generated was vital in Therefore, this benefit minimized the number of
explaining trends formulated from the analysis of users who clicked to obtain a result and ensured
quantitative data obtained from the baseline and consistent data processing.
transition phases. The exit survey was a standard
document provided by the MECS team and was Finally, the consultant developed the PowerPoint
utilized as is, with minor logical flow changes. presentation and provided the raw data for further
feedback from the client. The PowerPoint presen-
tation aimed to summarize the initial findings and
2.4 Synthesis and Reporting obtain further feedback for the final report.
Analysis of the data collected was governed by The client provided further feedback requiring
the cooking diaries (ANNEX 1) analysis guide. Even further data cleaning and alignment of this data
though the cooking diary is a generic document, with the final report.
its use was critical to answering the research
Figure 10: E-cooking training using the EPC and hotplate. Source:EED
14
03
Analysis and results
3.1 Household characteristics
Ten households located within Kawangware were gone to college, university, or vocational courses).
selected for the study. The households’ main Assessing the stoves owned per household yielded
cooks included nine (9) women and one (1) man. the stoves set in Figure 11, with their ownership dis-
The average household size across all households tribution represented in Table 5. It is noted that
was 4.4 members. Of these households, five (5) LPG cooking solutions were standard in all ten
were from the low-income category (spending households. Common electric cooking appliances
US$ 382 or less per month), and five were from (e.g., microwaves) were present in almost all
the middle-income category (spending above 382 households. Emerging electric appliances in Kenya,
USD, but below1,518 USD per month). Most of the such as rice cookers and EPCs, were present in up
main cooks had a tertiary level of education (had to two households.
Figure 11: Baseline stoves owned across households. Source:EED
Although the sample size for this study is small compared to the Kenya cooking sector (MoE, 2019), the
findings corroborate LPG fuel being the prevalent primary cooking solution in urban areas at 46%12. There
is also a diversification in the electric cooking appliances at the household level, as shown in Table 5, e.g.,
the presence of rice cookers and EPCs which were not identified in the cooking solution mix five years ago
(MoE,2019).
12. Ministry of Energy (2019). Kenya Household Cooking Sector Study: Assessment of the Supply and Demand of Cooking Solutions at the
Household Level
15
Table 5: Summary of stoves owned across all households as recorded in the entry survey
# Stoves type Total stoves found
1 Three-stone open-fire stove 1
2 Ceramic charcoal stoves 2
3 Improved Biomass Cookstove 1
4 Koko ethanol stove 2
5 LPG 6-kg cylinder with grill 8
6 LPG/electric cooker with electric grill/oven 3
7 LPG/electric cooker with gas grill/oven 2
8 LPG stove (tabletop stove) 2
9 Microwave 8
10 Rice cookers 2
11 Electric pressure cookers 1
12 Electric kettle 6
Nine of ten households completed the research As a result, all analyses detailed in the subsequent
study’s baseline and transition phases. Analysis of sections, except barrier analysis, utilize data from 8
the 9 (nine) household data showed data falsifica- households except section 3.4, which utilizes data
tion in one household and insufficient LPG data from from 7 (seven) households due to the absence of
a second household. The cause of the data falsifica- LPG weight measurements. A display of fuel use
tion in the former household is attributed to cooking across all eight households is shown in Table 6 and
by third parties and respondent fatigue. The insuf- Table 7 for the baseline and transition, respective-
ficient LPG data in the latter household is caused ly – households are identified using surname for
by the over-use of the 13-kg M-GAS LPG canister anonymity. Barrier analysis utilized data from all 8
by the household in the baseline, which yielded no (eight) households. The classifications of the final 8
weight measurements required for energy compar- (eight) households are displayed in Table 8.
isons, as measurements were recorded in Kenya
shillings as displayed by the M-GAS Pay go module.
Table 6: Frequency of use of cooking devices within the baseline phase
# Household by surname Air fryer Charcoal Oven Gas M-gas Microwave

stove stove
1 Kayumba - - - 42 - 16
2 Ondara - 5 - 24 - -
3 Abulitsa - 4 - 70 - 11
4 Waithaka - - - 77 6 33
5 Musyoki - - - 26 - 12
6 Njoki 2 - 7 15 - 7
7 Kamau - - - 33 - 8
8 Njoroge - - - - 39 -
16
Table 7: Frequency of use of cooking devices within the transition phase
# Household Air Charcoal Hotplate Oven EPC Gas stove M-gas Microwave
by surname fryer stove
1 Kayumba - - 23 - 10 - - 9
2 Ondara - - 15 - 18 11 - -
3 Abulitsa 5 - 25 - 35 10 - 11
4 Waithaka 4 - 9 - 35 2 3 45
5 Musyoki 5 1 14 - 23 14 - 20
6 Njoki 7 - 8 1 24 11 - 10
7 Kamau - - 6 - 11 4 - 10
8 Njoroge - - - - 14 - 7 -
Table 8: Classifications of households by income group
# Household ID Household by surname Income group

1 HH06 Kayumba Middle income
2 HH10 Ondara Low income
3 HH08 Abulitsa Low income
4 HH02 Waithaka Middle income
5 HH04 Musyoki Middle income
6 HH07 Njoki Low income
7 HH05 Kamau Middle income
8 HH01 Njoroge Middle income
3.2 Cooking diaries overview

3.2.1 Frequency of dishes and meals prepared during the study period
The consultant collected 898 data points over the as illustrated in Table 9. This observation can be
study period (498 data points during the baseline attributed to societal practices where children stay
phase and 460 data points during the transition temporarily with extended relatives during school
phase) from 8 (eight) households on individual holidays in December, away from their parents. In
dishes recorded over four (4) weeks. There was contrast, the decrease in adults is attributed to more
a reduction and increase (although minimal) in adult guests in the baseline and the resumption
the number of adults and children catered per of work activities in the transition phase (January
dish between the baseline (December 2022) and 2023).
transition (January 2023) phases, respectively,
Table 9: Mean number of adults and children cooked catered per dish across research phases
# Research phase Mean adults per dish Mean children per dish
1 Baseline 3.0 0.8
2 Transition 2.6 1.0
17
Understanding the more significant meals depends people catered, but this is more significant during
on how many times each meal type, such as the holidays, where a greater number of guests
breakfast, lunch, and supper, was prepared and may be received during lunch hour, in contrast to
the number of people catered per meal. Meal workdays. However, Table 10 shows that supper
types such as breakfast, lunch and supper is the most popular meal in both study phases,
comprise 96.2% of all meals designed, while 3.8% cementing supper’s constant importance across
accounted for others (e.g. baby food, snacks, etc.). holidays and workdays.
Data in Figure 12 shows that lunch meals have more
Table 10: Number of times each meal was cooked per research phase
# Meal type Frequency in baseline (N) Frequency in transition (N)

1 Supper 243 266
2 Lunch 106 117
3 Breakfast 69 63
4 Others (baby food, snacks, etc.) 20 14
Figure 12: Mean number of people catered per meal, classified by study phase
3
Mean number of people catered per meal
0
Breakfast Lunch Supper
Meal type
Phase Baseline Transition
18
3.2.2 Dish quantity per meal
Despite the slightly lower number of

average people catered per meal in the During the transition period, single-dish meals
transition period, as detailed in Figure 12, decreased in popularity.
households showed a greater propensity During the transition period, more complex
to create more complex meals using meals comprising three (3), four (4) and five (5)
electric cooking, as described in Figure dishes increased in popularity.
13, with the following general differenc-
es noted between the baseline and During the transition period, meals containing
transition periods: two (2) dishes showed a marginal decrease in
popularity.
Figure 13: Average trends in meal complexity by the number of dishes prepared per meal
50%
40%
30%
Percentage of meals
20%
10%
0%
1 2 3 4 5
Dishes per meal
19
The increase in meal complexity may be attributed The simplest meal, by the number of dishes
to a combination of factors such as ease of EPC use cooked, was breakfast, whereas the most complex
and the positive effect of the electricity subsidy en- meal was supper. Interestingly, most breakfast and
couraging greater use of electricity but conversely lunch meals have one dish. In contrast, most supper
leading to an unintended increase in the number meals have two dishes prepared and are the most
of dishes cooked per meal. consistently cooked across workdays and holidays.
Due to most supper meals comprising two dishes,
We further aggregated the data by meal to visualize supper is the most complex meal. This data further
the complexity of each meal type. This complexity agrees with social norms where breakfast and
across meals, is shown in Figure 14, Figure 15 and lunch meals are simpler with more people present
Figure 16. It is deducible that meal complexity during supper.
increases with the time a meal is prepared.
Figure 14: Complexity of breakfast meals by the number of dishes prepared per meal
70%
60%
Percentage of Breakfast meals
50%
40%
30%
20%
10%
0%
1 2 3 4
Dish quantity
20
Figure 15: Complexity of lunch meals by the number of dishes prepared per meal
60%
50%
40%
Percentage of lunch meals
30%
20%
10%
0%
1 2 3 4 5
Dish quantity
Figure 16: Complexity of supper meals by the number of dishes prepared per meal
35%
30%
25%
Percentage of lsupper meals
20%
15%
10%
5%
0%
1 2 3 4 5
Dish quantity
21
3.2.3 Changes in dish state

Three dish states are defined by the cooking diaries Table 11: Mean number of adults and children cooked
catered per dish across research phases
protocol 3.0 – Fresh dishes, reheated dishes, and
partially pre-cooked dishes. The consultant noted
# Research phase Baseline Transition
that households mainly cooked fresh dishes. Such
dishes have the most significant total dishes at 1 Fresh dish 281 251
64.1% in the baseline and 54.5% in the transition 2 Reheated dish 127 151
phase. Reheated dishes had the second highest 3 Partially cooked 30 58
significance in the study, accounting for 30.0% and dish
32.8% in the baseline and transition phases, re-
spectively. Partially cooked dishes represented
the least occurring category within the study. The
general dish classification across the study is listed
in Table 11.
©Shutterstock
22
Figure 17 shows that fresh dishes are the most meals. However, in the transition phase, reheated
popular type within supper meals in both the meals were most common. This change could be
baseline and the transition. Reheated dishes attributed to the resumption of work activities in the
represent the second most popular dish state, and transition phase (January 2023) compared to the
partially precooked meals are the least popular. holiday period (December 2022) that coincided
In Figure 18 and Figure 19, It is shown that there with the baseline phase. Such work activities result
is a change in the most popular dish state across in less time in the kitchen, causing an increase in
lunch and breakfast meals. Fresh meals are popular reheated meals.
in the baseline phase for lunch and breakfast
Figure 17: Changes in dish state during supper meals across baseline and transition phases
160
140
Number of supper dishes (N)
120
100
80
60
40
20
0
Baseline Transition
Phase
Fresh Partially precooked Reheated
Figure 18: Changes in dish state during lunch meals across baseline and transition phases
60
50
Number of supper dishes (N)
40
30
20
10
0
Baseline Transition
Phase
23
Figure 19: Changes in dish state during breakfast meals across baseline and transition phases
40
Number of breakfast dishes (N)
30
20
10
0
Baseline Transition
Phase
3.2.4 Fuel popularity per meal

For both research phases, households utilized an Analysis shows that charcoal is used more at 8.5%
assortment of fuels in varying proportions. During of all dishes cooked during the baseline than 0.9%
the baseline phase, LPG was the most used fuel, in the transition. This change is attributed to the
with a usage of 76.0%, followed by electricity at more significant number of people catered during
21.9% and charcoal at 2.1%. Most of the electricity lunch meals in the baseline in December 2022
usage in the baseline was mainly from reheating compared to the transition in January 2023 (Figure
using microwaves and not actual cooking using 12). In the transition, charcoal use is attributed to
other electric cooking devices. In the transition precook beans as its only use (Figure 25). The
phase, electricity was the most used fuel, with a usage of charcoal for lunch meals in the baseline
usage ratio of 86.3%. The utilization of LPG dropped may point to a more significant number of people
to 13.5%, and charcoal use went to as low as 0.2% catered per dish, where most baseline lunch meals
for all the prepared dishes. It is evident that despite involve a single dish (Figure 15) to serve a greater
the motivation to use electricity, households still number of people compared to the transition phase
cooked using LPG and charcoal, necessitating a (Figure 12). More detail on fuel use per meal is
further analysis of the foods cooked using these provided in Table 12, Table 13 and Table 14.
fuels in both phases of the study.
Table 12: Fuel use by percentage during supper meals across baseline and transition phases
# Fuel Baseline supper dishes (%) Transition supper dishes (%)
1 LPG 85.2% 14.7%
2 Electricity 14.8% 85.3%
24
Table 13: Fuel use by percentage during lunch meals across baseline and transition phases
# Fuel Baseline lunch dishes (%) Transition lunch dishes (%)

1 LPG 63.2% 12.8%
3 Charcoal 8.5% 0.9%
Table 14: Fuel use by percentage during breakfast meals across baseline and transition phases
# Fuel Baseline breakfast dishes (%) Transition breakfast dishes (%)
1 LPG 63.8% 7.9%
3.2.5 Dishes cooked vs. type of fuel

The households cooked 46 (forty-six) unique dishes and leafy green vegetables and tubers. However,
in each phase – the baseline and transition phases. despite the training provided in the transition
LPG, the leading cooking solution during the phase, the local cuisine-friendly interface of the
baseline phase, was used in most dishes cooked EPC that contains ugali among other foods, and
in the baseline. Due to the large variety of dishes an e-cookbook guide provided per household –
cooked, we limit our focus to the subset of the top ugali dishes occupy 19.7% of all dishes cooked on
ten dishes. Of this subset, a bar chart is visualized the LPG in the transition, compared to a usage of
in Figure 20 and Figure 21, which show ugali and 14.8% in the baseline, pointing to more extensive
rice as the most popular meals in the baseline, efforts required in training and behavioural change.
followed by rice, beans, an array of meaty dishes,
Figure 20: Percentage usage of LPG by dish in the baseline
14%
Percentage of heating events using LPG
12%
10%
8%
6%
4%
2%
0%
Ugali Rice/ Beans/ Fried Beef/goat/ Spinach/ Tea Cabbage Mrenda/ Chicken/
Mchele Maharagwe eggs mutton/ sukuma Mboga Duck
rabbit wiki/ lettuce/ kienyeji
broccoli
LPG Energy source
25
Figure 21: Percentage usage of LPG by dish in the transition

20%
18%
Percentage of heating events using LPG
15%
12%
10%
8%
5%
2%
0%
Ugali Chapati Beef/goat/ Fried Rice/ Pasta/ Beans/ Cabbage French Spinach/
mutton/ eggs Mchele Sphagetti Maharagwe fries/ sukuma
rabbit chips/ wiki/ lettuce/
bhajia broccoli
LPG Energy source
The consultant noted that the baseline electricity Despite chapati being the next popular dish, par-
usage in Figure 22 was mainly from the microwave ticipants were not able to complete their chapatti
and, particularly, baking in one household. Of all (flat bread) cooking activities due to the long time
foods cooked in the transition, as shown in Figure required to heat the cooking utensil with the electric
23 using the EPC, hotplate, and air fryer – the rice hotplate, which caused drying of the chapattis while
was the most popular dish, with several meaty cooking, causing the electric hotplate to be incom-
dishes being easier to prepare. The electricity patible with the cooking processes for chapatti
usage shows that rice was the most popular dish dishes.
on EPCs.
Figure 22: Percentage usage of electricity by dish in the baseline
30%
Percentage of heating events using Electricity
25%
20%
15%
10%
5%
0%
Rice/ Chapati Beans/ Sweet Chicken/ Ugali Potatoes Milk Beef/goat/ Cake
Mchele Maharagwe potatoes/ Duck mutton/
Arrow roots/ rabbit
Cassava/
Ngwacii/
Nduma/
Muhogo
Electricity Energy source

26
Figure 23: Percentage usage of electricity by dish in the transition
16%
14%
12%
10%
8%
6%
4%
2%
0%
Rice/ Ugali Beef/goat/ Sweet Chicken/ Beans/ Spinach/ Mrenda/ Green Cabbage
Mchele mutton/ potatoes/ Duck Maharagwe sukuma Mboga grams/
rabbit Arrow roots/ wiki/ lettuce/ kienyeji Ndengu
Cassava/ broccoli
Ngwacii/
Nduma/
Muhogo
Figure 24 shows that charcoal is mainly used cooking solution (frying precooked grains on an
for cooking meats and chapatti in the baseline. LPG stove). The consultant also noted that beans
However, it was noted that the pre-cooking of large are not present as one of the dishes cooked using
batches of cereals was done using charcoal, while charcoal in the baseline, as shown in Figure 24; it is
small batches of pre-cooked grains were prepared the only use of charcoal in the transition, as shown
using alternative fuels such as LPG. This observa- in Figure 25. A possible reason for this difference
tion further explains the fuel stacking phenomena in cereal preparation behaviour using charcoal
observed in a household where a household between the baseline and transition phases is due
prefers one type of cooking solution for a specific to the rural-to-city migration after the December
cooking activity (e.g., boiling cereals on a charcoal holiday, which led to batch cooking and cereals’
stove) and another cooking activity on a different storage during the transition phase in January.
©Shutterstock
27
Figure 24: Percentage usage of charcoal by dish in the baseline
30%
25%
Percentage of heating events using Charcoal
20%
15%
10%
5%
0%
Beef/goat/ Chicken/ Chapatti Matumbo Nyama Pilau
mutton/ Duck choma
rabbit
Charcoal Energy source
Figure 25: Percentage usage of charcoal by dish in the 3.2.6 Dishes cooked Vs. Cooking device
transition
Across the households, Table 15 shows the utiliza-
100% tion of cooking devices within the study. The most
popular among households was the 6-kg single
burner stove, with a 65.8% usage across all dishes,
Percentage of heating events using Charcoal
with microwaves being the second most popular

80% cooking device. One notable device is M-gas
(Circle Gas, London, United Kingdom), a pay-as-
you-cook LPG device that was present in two (2)
60% households out of the 9 (nine) households.
In the transition phase, the three most utilized

devices were the EPC, microwave, and hotplate.
40% The air fryer was supplied to only four (4)
households whose electrical wiring could withstand
the electrical load introduced by three parallel
20% cooking appliances – the EPC, the hotplate, and
the air fryer.
0%
Beans/ Maharagwe
Charcoal Energy source
28
Table 15: Overall usage of cooking devices across dishes
Fuel Cooking device Baseline utilization Transition utilization

LPG 6-kg single burner stove 65.8% 11.3%
M-gas double burner 10.3% 2.2%
Electricity EPC - 37.0%
Electric hotplate - 21.7%
Air fryer 0.5% 4.6%
Microwave 19.9% 22.8%
Electric oven 1.6% 0.2%
Charcoal Charcoal stove 2.1% 0.2%
The consultant performed further analysis to cooking appliances. This attribution is compounded
determine the most popular meals cooked on each by the analysis in Figure 16, which shows that some
appliance during each study phase to determine meals, such as supper, contain mainly two dishes
the impact of an electric transition on the usage causing the households to prefer to cook ugali
of the appliances. The gas stove, the most utilized on the gas stove and utilize the EPC for an ac-
appliance in the baseline, was used for various companying dish. An additional notable change
dishes, with Ugali as the most frequently cooked is an increase in cooking chapatti on the gas
dish on the gas stove. Despite the reduction in gas stove between the baseline in Figure 26 and the
stove use in the transition at 11.3% compared to transition in Figure 27 due to the inefficient heating
65.8% in the transition - ugali gains a larger share on the hotplate, which caused chapattis (flat breads)
of use on the gas stove in the transition with an to dry instead of cook. This increase in cooking
8% lead on rice that is the secondly most cooked chapatti is also present in the M-gas double burner
dish on the gas stove, compare d to a 1% lead in cookers, as shown in the difference between the
the baseline. This change may be attributed to M-gas dishes in the baseline in Figure 28 and the
households not shifting ugali cooking to electric transition phase in Figure 29.
Figure 26: Most frequently cooked dishes on the gas stove (6 kg) in the baseline phase
14%
12%
Gas stove utilization during Baseline (%)
10%
8%
6%
4%
2%
0%
Ugali Rice/ Beans/ Beef/goat/ Spinach/ Tea Fried Chicken/ Cabbage Chapati
Mchele Maharagwe mutton/ sukuma eggs Duck
rabbit wiki/ lettuce/
broccoli
29
Figure 27: Most frequently cooked dishes on the gas stove (6 kg) in the transition phase
20%
18%
15%
12%
10%
8%
5%
2%
0%
Ugali Rice/ Beans/ Pasta/ Beef/goat/ Chapati Fried French Spinach/ Cabbage
Mchele Maharagwe Sphagetti mutton/ eggs fries/ Chips/ sukuma
rabbit Bhajia wiki/ lettuce/
broccoli
Gas stove Cooking device
Figure 28: Most frequently cooked dishes on the M-gas double burner gas stove in the baseline phase
18%
15%
Mgas utilization during Baseline (%)
12%
10%
8%
5%
2%
0%
Fried Rice/ Ugali Beans/ Matoke Spinach/ Githeri Green Mrenda/ Cabbage
eggs Mchele Maharagwe sukuma grams/ Mboga
wiki/ lettuce/ Ndengu kienyeji
broccoli
Mgas Cooking device
30
Figure 29: Most frequently cooked dishes on the M-gas double burner gas stove in the transition phase
20%
18%
15%
Mgas utilization during Transition (%)
12%
10%
8%
5%
2%
0%
Chapati Fried Ugali Cabbage Cow Porridge/ Beef/goat/
eggs peas oats/ mutton/
granola rabbit
Mgas Cooking device
As the charcoal stove was the only charcoal fuel transition phase occurred. The resumption of work
cooking device, dishes cooked using the charcoal may be correlated with a corresponding decrease
stove in Figure 30 and Figure 31 are identical to in the time available to cook fresh dishes and an
Figure 24 and Figure 25, showing the dishes increase in reheating of pre-cooked dishes.
cooked using charcoal as a fuel. Charcoal usage
data from these figures show the reduced use of EPCs and hotplates were utilized during the
the charcoal stove in the transition phase. This transition phase alone. Meats and rice were the
reduction may be interpreted that given suitable most frequently cooked meals using the EPC.
external factors promoting other fuels, such as Further analysis shows that rice was the most
cost, and availability of the right electric appliances, prepared using the electric hotplate. This utili-
households can shift away from charcoal. sation is illustrated in Figure 34 and Figure 37.
A few households indicated that they liked the
Microwaves were primarily used to reheat dishes. taste of rice cooked on the electric hotplate. This
Rice was the most reheated dish in both the perception may be attributed to the slow cooking
baseline and transition phases. This use is illus- of the hotplate, resulting in well-cooked rice. The
trated in Figure 32 and Figure 33. In addition to air fryer was used in one household to cook two
the increased use of microwaves in the transition specific dishes – Mandazi, a Swahili fried bread,
phase, we noted increased reheating of non-rice and potatoes. The increased use of the air fryer in
meals. This increase may be attributed to the rise the transition yielded an increase in dish varieties
of reheating dishes during lunch and breakfast, prepared on the air fryer from two to six, as noted
as shown in Figure 18 and Figure 19, due to the in Figure 35 and Figure 36.
resumption of work activities in January when the
31
Figure 30: Most frequently cooked dishes on the charcoal stove in the baseline phase
30%
25%
Charcoal stove utilization during Baseline (%)
20%
15%
10%
5%
0%
Beef/goat/ Chicken/ Chapati Matumbo Nyama Pilau
mutton/ Duck choma
rabbit
Charcoal stove Cooking Device
Figure 31: Most frequently cooked dishes on the

charcoal stove in the transition phase
100%
Charcoal stove utilization during Transition (%)
80%
60%
40%
20%
0%
Beans/ Maharagwe
Charcoal stove
Cooking device
©EED
32
Figure 32: Most frequently reheated dishes on the microwave in the baseline phase
35%
Microwave utilization during Baseline (%)
30%
25%
20%
15%
10%
5%
0%
Rice/ Chapati Beans/ Sweet Chicken/ Ugali Milk Beef/goat/ Matoke Pilau
Mchele Maharagwe Potatoes/ Duck mutton/
Arrow Roots/ rabbit
Cassava/
Ngwacii/
Nduma/
Muhogo
Microwave cooking device
Figure 33: Most frequently reheated dishes on the microwave in the transition phase
30%
Microwave utilization during Transition (%)
25%
20%
15%
10%
5%
0%
Rice/ Ugali Sweet Chapati Beef/ Goat/ Cow French Fries/ Cabbage Mukimo Pizza
Mchele Potatoes/ Mutton/ Peas Chips/ Bhajia
Arrow Roots/ Rabbit
Cassava/
Ngwacii/
Nduma/
Muhogo
Microwave cooking device
33
Figure 34: Most frequently cooked dishes on the EPC in the transition phase
14%
Electric pressure cooker utilization during Transition (%)
12%
10%
8%
6%
4%
2%
0%
Beef/ Goat/ Rice/ Ugali Bean/ Chicken/ Green Sweet Spinach/ Mrenda/ Matoke
Mutton/ Mchele maharagwe Duck grams/ Potatoes/ Arrow Sukuma Mboga
Rabbit Ndengu Roots/ Cassava/ Wiki/ Kienyeji
Ngwacii/ Lettuce/
Nduma/ Broccoli
Muhogo
Electric pressure cooker Cooking device
Figure 35: Most frequently cooked dishes on the air fryer in the
baseline phase
50%
Air fryer utilization during Baseline (%)
40%
30%
20%
10%
0%
Maandazi/ Bhajia
Mahamri ©EED
Air fryer cooking device
34
Figure 36: Most frequently cooked dishes on the air fryer in the transition phase
40%
35%
Air fryer utilization during Transition (%)
30%
25%
20%
15%
10%
5%
0%
Sausages French Fries/ Sweet Potatoes/ Beef/ Goat/ Njugu Fish/ Tilapia/
Chips/ Bhajia Arrow Roots/ Mutton/ Samaki/ Fillet
Cassava/ Ngwacii/ Rabbit
Nduma/ Muhogo
Air fryer cooking device
Figure 37: Most frequently cooked dishes on the electric hotplate in the transition phase
12%
Electric hotplate utilization during Transition (%)
10%
8%
6%
4%
2%
0%
Rice/ Mrenda/ Ugali Spinach/ Fried Pasta/ Beef/ Goat/ Fish/ Chicken/ Porridge/
Mchele Mboga Sukuma Wiki/ Eggs Spaghetti Mutton/ Tilapia/ Duck Oats/
Kienyeji Lettuce/ Rabbit Samaki/ Granola
Broccoli Fillet
Electric hotplate cooking device
35
3.3 Time, energy, and cost savings

3.3.1 Energy per capita and cook time per dish
The consultant split the dishes prepared into profiles. In addition to these analyses, Table 16 and
the three dish state categories – fresh, partially Table 17 show the cooking time and energy use
cooked, and reheated. The cooking time was for fresh dishes, disaggregated by the cooking
analysed across the major dish state categories, device used. In contrast, Table 20 and Table 21
such as fresh and reheated dishes. Partially cooked show the cook time and energy use for reheated
dishes were, however, not analysed due to a lack dishes, disaggregated by the cooking device used,
of enough data. The consultant only compared which allows fine-grained visibility of the energy
electric cooking to LPG due to insufficient data on consumption and cooking time across the cooking
charcoal, leading to incorrect descriptive statistics devices, in addition to the fuel used.
such as the data mean.
Table 16 shows that the long cook times on a
Table 16, Table 17 and Table 18 show the cooking hotplate and the pre-programmed cooking time
time and energy use for fresh dishes, disaggregat- on the EPC and air fryer may not save time for
ed by the fuel used. In contrast, Table 19 , Table 20 cooking fresh dishes. Table 19 shows that electric
and Table 21 show the cooking time and energy use cooking does, however, yield time savings for
for reheated dishes. It is worth noting that cooking reheated dishes due to the faster reheating time
using the EPC and air fryer was performed utilizing while utilizing the microwave. However, despite
the set cook times on the appliances. As such, the the cooking time concerns, the automated cooking
recorded time may not represent the actual cook on an EPC does yield a simpler cooking process as
time for these appliances as it is not entirely under the EPC can be left unattended. However, electric
user control due to limited user knowledge of when cooking utilises less energy for fresh and reheated
to change the default cooking time and heating dishes, as shown in Table 18 and Table 21.
Table 16: Cooking time and energy used in cooking fresh dishes
# Dish prepared LPG energy per Electricity energy LPG cook- Electricity cook-
capita (MJ) per capita (MJ) time (mins) time (mins)
1 Beans / Maharagwe 19.7 2.0 78.7 73.8
2 Boiled eggs 5.5 0.8 21.3 18.0
3 Cabbage 2.1 0.7 16.9 18.0
4 Chicken/duck 6.3 1.8 76.6 50.1
5 French fries/chips/Bhajia 5.0 1.7 40.9 44.1
6 Fried eggs 2.5 1.8 11.8 33.3
7 Matoke 2.1 2.1 38.0 39.4
8 Mrenda/ Mboga kienyeji 1.5 2.4 20.9 51.6
9 Porridge/oats/granola 10.8 1.2 37.8 32.0
10 Potatoes 4.5 1.3 37.5 41.2
11 Rice / Mchele 4.5 1.5 40.6 38.1
12 Spinach 4.5 1.0 24.0 33.2
13 Sweet potatoes /Tubers 8.9 1.0 52.4 26.7
14 Ugali 4.6 1.7 35.9 57.0
15 beef/goat/Mutton/Rabbit 9.0 1.8 47.4 38.8
36
Table 17: Cooking time for various fresh dishes across cooking devices in minutes
# Dish prepared Air fryer Hotplate Oven EPC Gas stove
1 Beans / Maharagwe - - - 73.8 81.4
2 Boiled eggs - - - 18.0 21.3
3 Cabbage - 17.7 - 18.2 -
4 Chicken/duck - - - 54.3 76.6
5 French fries/chips 35.8 42.5 - 81.0 40.9
6 Fried eggs - 33.3 - - 13.3
7 Matoke - - - 45.7 33.6
8 Mrenda/ Mboga kienyeji - 23.0 - 56.3 20.7
9 Porridge/oats/granola - 32.0 - - 43.4
10 Potatoes - 32.0 31.0 - 37.5
11 Rice / Mchele - - - 36.8 39.3
12 Spinach - - - 30.0 26.7
13 Sweet potatoes /Tuber - - - 26.7 52.4
14 Ugali - 38.5 - 63.9 37.2
15 beef/goat / Mutton 18.0 31.0 - 44.4 50.4
Table 18: Fresh dish energy per capita requirements for cooking across various cooking devices in Megajoules
# Dish prepared Air fryer Hotplate Oven EPC Gas stove
1 Beans / Maharagwe - - - 2.0 19.7
2 Boiled eggs - - 0.8 5.5
3 Cabbage - 0.5 0.9 -
4 Chicken/duck - 1.8 6.3
5 French fries/chips 2.2 1.3 - 0.8 6.3
6 Fried eggs - 1.8 - - 2.5
7 Matoke - - - 2.1 2.1
8 Mrenda/ Mboga kienyeji - 0.9 - 2.6 1.5
9 Porridge/oats/granola - 1.2 - - 10.8
10 Potatoes - 0.9 1.8 - 4.5
11 Rice / Mchele - - - 1.6 4.5
12 Spinach - - 0.9 4.5
13 Sweet potatoes /Tuber - - - 1.0 8.9
14 Ugali 1.4 - 1.9 4.6
15 beef/goat / Mutton 0.9 0.1 - 2.2 9.0
37
Table 19: Cooking time for various reheated dishes across cooking fuels in minutes
# Phase LPG energy per Electricity LPG cook-time Electricity cook-

capita (MJ)) energy per (mins) time (mins)
capita (MJ)
1 Beans 2.7 0.7 10.9 4.9
2 Rice 6.8 0.3 15.9 5.2
3 Beef/goat/Mutton/
1.8 0.6 9.4 10.1
Rabbit
Table 20: Cooking time for various reheated dishes across cooking devices in minutes
# Dish prepared Air Hotplate EPC Gas stove M-gas burner Microwave
fryer
1 Beans - 20.0 11.0 10.7 13.0 2.6

2 Rice - 13.2 6.0 19.2 9.3 4.6
Beef/goat/mutton/ 19.7 13.0 9.4 5.6
3 10.0 -
rabbit
Table 21: Reheated dish energy per capita requirements for cooking across various fuels in Megajoules
# Phase Air fryer Hotplate Oven EPC Gas stove Microwave
1 Beans - - - 2.0 19.7
2 Rice - 1.2 - 1.6 4.5
3 Beef/goat / Mutton / Rabbit 0.9 0.1 2.2 9.0 0.4
3.3.2 Energy per capita and cook time per meal
As a meal comprises dishes served together, the parallel use of LPG (especially those with multiple
consultant further assessed the cooking time per burners), charcoal, and electric appliances. Ad-
meal, such as breakfast, lunch and supper. In Table ditionally, the difficulty associated with emptying
22 below, during breakfast and lunch, there were the first dish and cleaning the cooking pot to start
reductions in cooking time at the meal level of 47.7% cooking the second dish in the case of EPCs
and 40%, respectively, due to the presence of dish explains the increase in cooking times, especially
reheating utilizing microwaves in the transition, as for supper meals where more dishes are cooked,
visualized in Figure 17 and Figure 18. This reduction as discussed earlier.
could be attributed to faster lunch and meal
cooking times to the resumption of work activities
in the transition (January 2023), compared to the
baseline (December 2022) that coincided with the
December holiday period.
This evidence may be further compounded by

the study’s constraint, where each household
only received a single EPC, hotplate and possibly
an additional air fryer. Due to the slow cooking
times experienced with the hotplate, participants
could not parallel cook effectively using electric
appliances. However, despite the increase in
supper cooking times, the increase is only 20.1%
of the baseline. This increase is attributed to the
©EED
38
Table 22: Mean cook time per meal
# Phase Mean supper cook time Mean lunch cook time Mean breakfast cook time
1 Baseline 30.4 minutes 35.5 minutes 27.0 minutes
2 Transition 36.5 minutes 21.3 minutes 14.1 minutes
Analysis of the energy per meal is shown in Table 23. The data shows that the baseline’s most energy-inten-
sive meal by average measure is breakfast, while lunch meals are in the transition. The consultant obtained
higher average energy than the baseline’s median (50%) due to the higher use of LPG and charcoal, as
shown in Table 13.
Table 23: Per capita energy requirement statistics per meal across baseline and transition
# Phase Meal N mean Std dev 25% 50% 75 %

1 Baseline Breakfast 30 2.61 3.54 0.03 0.58 0.97
2 Baseline Lunch 72 2.2 3.04 0.07 0.46 0.84
3 Baseline Snack 12 2.31 3.71 0.28 0.69 1.03
4 Baseline Supper 153 1.17 1.22 0.01 0.46 0.77
5 Transition Breakfast 31 0.42 0.47 0.01 0.14 0.29
6 Transition Lunch 69 1.11 4.57 0.01 0.16 0.31
7 Transition Snack 11 0.62 0.68 0.06 0.10 0.27
8 Transition Supper 212 0.83 2.68 0.01 0.19 0.36
3.3.3 Energy per capita per day

Analysis of the total energy per capita per day in the baseline in Table 24 shows a larger median
per capita energy across three households - Abulitsa’s, Waithaka’s and Musyoki’s. The larger
energy use across these households may be attributed to poorer cooking styles resulting in
lower cooking efficiencies. However, the consultant could not attribute this to the number of
people in the households or the stoves in Table 24.
Table 24: Baseline energy per capita per day

# Household ID N mean std 25% 50% 75%
1 Kayumba 17 1.31 1.39 0.00 1.15 2.15
2 Ondara 12 4.64 10.11 0.99 1.59 2.67
3 Abulitsa 18 7.76 9.64 2.13 3.8 12.78
4 Waithaka 16 5.78 5.39 2.66 4.04 7.43
5 Musyoki 12 5.67 6.78 1.21 3.8 7.13
6 Njoki 13 2.88 4.4 0.00 0.77 4.92
7 Kamau 15 1.86 1.85 0.64 1.47 2.3
Table 25 details the median per capita energy per day within the transition across fuels used,
including LPG, charcoal and electricity. The higher energy use shown by a larger median per
capita energy is attributed to the higher use of LPG and charcoal across select households within
the transition phase. These households include Ondara’s, Musyoki’s and Njoki’s households.
39
Table 25: Transition energy per capita per day

# Household ID N mean std 25% 50% 75%
1 Kayumba 15 1.08 0.74 0.78 1 1.48
2 Ondara 14 2.55 2.24 0.67 1.67 4.6
3 Abulitsa 16 2.73 2.48 0.71 2.21 3.94
4 Waithaka 15 1.36 2.39 0.4 0.65 0.98
5 Musyoki 14 7.41 14.05 1.02 2.16 3.48
6 Njoki 14 2.41 2.13 1.02 1.99 3.59
7 Kamau 12 1.53 2.1 0.29 0.55 2.24
3.3.4 Energy and cost dynamics

The cooking diaries protocol compares the energy We, therefore, calculate the level of electric
dynamics of various standard fuels, such as LPG transition using the equation below:
or charcoal, with electricity. A primary part of the
cooking diaries protocol is the unstacked baseline The distribution of household energy costs per
phase which requires collecting numerous heating person per dish and household energy used per
data points with the baseline fuels to collect person per dish are tabulated in Table 26 and Table
adequate data to allow comparisons between the 27 for the baseline and transition phases, respec-
baseline fuels and electricity. In the baseline, the tively. It is evident that the greater the proportion of
primary baseline fuels across all households were dishes prepared with electricity, the lower the per
LPG and charcoal. Still, despite this, the consultant capita cost per dish. The decrease in dish energy
noted the presence of electric appliances, namely and dish costs with increasing electric transition
microwaves, EPCs, air fryers and rice cookers. is attributed to the greater use of efficient electric
cooking appliances such as the EPC.
An objective metric to measure transition is needed
to determine the implication of transitioning to
electric cooking.
Electric transition level (ŋ)= ( Total dishes cooked using electricity

Total dishes in research phase ) X 100% ,0≤ ŋ≤100
Table 26: Summary statistics on energy and cost dynamics of the baseline phase (using all baseline fuels – electricity,
LPG, and charcoal)
# Metric Percentage of HH dishes HH cost/person/dish HH energy/person/dish

(%) (US$) (MJ)
1 Mean 24.46 0.06 1.73
2 Standard deviation 16.20 0.03 0.68
3 Minimum 0.00 0.03 0.74
4 25th percentile 15.99 0.05 1.21
5 50th percentile 27.59 0.05 2.00
6 75th percentile 30.01 0.08 2.17
7 Maximum 51.61 0.10 2.62
40
Table 27: Summary statistics on energy and cost dynamics of the transition phase (using all transition fuels – electrici-
ty, LPG, and charcoal)
# Metric Percentage of HH dishes (%) HH cost/person/ HH energy/person/dish

dish (US$) (MJ)
1 Mean 86.84 0.03 0.84
2 Standard deviation 8.61 0.02 0.50
3 Minimum 75.00 0.02 0.37
4 25th percentile 81.24 0.03 0.55
5 50th percentile 87.10 0.03 0.75
6 75th percentile 91.64 0.04 0.89
7 Maximum 100.00 0.07 1.89
To determine the impact of a transition, we model the transition data using a straight-line model as shown
both inFigure 38, Figure 39 and, Figure 40, Figure 41 for energy and cost, respectively, vis-à-vis the electric
transition. From the model fit process using linear least squares fitting, the energy and cost model, per dish,
for an electric transition among the select households is as below:
Per capita energy per dish (MJ)= -0.0327η+3.6814

Per capita cost per dish (USD)= -0.0011η+0.1325
Figure 38: Baseline energy per capita required per dish against the level of transition.
2.50
2.25
2.00
Energy per capita (MJ)
1.75
1.50
1.25
1.00
0.75
0 10 20 30 40 50
Percentage of cooking events using electriicty (%)

Income goup
Middle income HH
Low income HH
41
Figure 39: Transition energy per capita required per dish against the level of transition
2.0
1.5
Energy per capita (MJ)
1.0
0.5
0.0
75 80 85 90 95 100
y= -0.0327x+3.6814
Figure 40: Baseline cost per capita per dish against the level of transition
0.10
0.09
Cost per capita (USD)
0.08
0.07
0.06
0.05
0.04
0.03
0 10 20 30 40 50
Income goup
Middle income HH
Low income HH
42
Figure 41: Transition cost per capita per dish against the level of transition
0.08
0.07
0.06
Cost per Capita (USD)
0.05
0.04
0.03
0.02
0.01
0.00
75 80 85 90 95 100

y= -0.0011x+0.1325
The models above on the energy and cost of 3.68MJ per dish, while 100% e-cook results
dynamics, therefore, provide answers to research in a per capita energy requirement of 0.41MJ
questions about an electric transition as follows: per dish. A household would save an average
of 3.27 MJ per capita.
No electric cooking within the study group
would result in a per capita cost of 0.13 US$ per With an energy cost of 0.11kWh (0.41MJ) per person
dish, while 100% e-cook results in a financial cost per dish, while cooking entirely with electricity, a
per capita of 0.02 US$ per dish. A household household with 2.9 household members, such as
would have an average cost saving per dish of those in Nairobi (KNBS, 2019), that cooks one to two
0.11 US$ per capita per dish cooked. dishes per meal daily across breakfast, lunch, and
supper would consume between 29.82 and 59.65
No electric cooking within the study group kWh of electricity monthly.
would result in a per capita energy requirement
29.82 - 59.65 kWh

The average monthly
electrical energy required for
cooking by a household with
2.9 members, such as those
in Nairobi (KNBS, 2019),
which cooks one to two
dishes per meal entirely with
electricity, across breakfast,
lunch, and supper meals.
©Shutterstock
43
04
End-user perceptions and
barriers to cooking entirely
with electricity
Based on the exit survey results, the consultant
obtained opinions on cooking experience
indicators such as cooking time, ease of heat
control, et cetera, while using various electric
cooking appliances. The data collected from the
exit survey was complemented by user stories
collected in an FGD conducted at the end of the
study period. The results from both activities allowed
the consultant to paint a clear picture of the use of
the provided electric cooking appliances, namely
EPCs, hotplates and air fryers. The consultant notes
that whereas the EPC and hotplates were distrib-
uted to all 9 (nine) households, the air fryers were
distributed to only four of the nine households
whose household electrical cabling could sustain
the parallel use of all 3 (three) appliances.
In addition to general user perception, each

household provided information on reasons for
not cooking using electricity at each meal using
the cooking diary records. The consultant compiled
and grouped these barriers, as shown in Table 28.
©SHUTTERSTOCK
Table 28: Summary of barriers
Category Specific
Supply of electricity Unreliable supply of electricity (frequent black-outs)
Shared electricity meters in some rental houses
Electric-cooking Slow cooking time for the hotplate case
appliance limitations
Inability to cater for large quantities of food
Inability to cook multiple dishes simultaneously on one appliance
Requirement for a specific cooking pot for the EPCs
Inability to cater for round/large pots for the case for a hotplate
Safety concerns for the hotplate
Complex procedures for a small dish
Incompatibility with some local dishes, e.g. chapatis for EPCs
User-Limitations Lack of awareness of the financial benefits of cooking entirely with electricity
Beliefs that certain meals are better cooked on a specific cooking solution, e.g.
ugali on LPG stoves
User preference on a specific cooking solution, e.g. LPG
44
Table 29 also further shows each barrier’s relative priority (as reported by the study participants from the
transition phase cooking diaries). This section elaborates further on barriers and user perception of cooking
entirely with electricity.
Table 29: Frequency of barriers encountered

# Reason Frequency (N)
1 Hotplate too slow 20
2 Blackout 17
3 Too much to cook 10
4 No prepaid energy balance 7
5 Too little to be cooked 6
6 LPG cooks better 5
7 Cook forgot 4
8 Cook was in a hurry 3
9 Inability to use electric appliances to cook 2
10 The air fryer takes smaller portions, hence moved to the gas 1
11 Lack of adult supervision 1
4.1 Assessment of e-cooking experience indicators

Three cooking experience indicators, chosen and the hotplate receiving 80% approval across
from the exit survey, for assessment of e-cook those households that received each appliance
appliances include: type. The consultant, however, recognizes that
the ability to control the heat is independent of
Heat regulation capability, the metrics comparing if the e-cook appliances
provide enough heating, which may affect
Satisfaction with the cooking pots used on the cooking speed.
e-cook appliance,
The households were asked about the cooking
General cooking speed across regular and speed while using various e-cook appliances.
long-cook dishes, The highest approval ratings were conferred
to the EPC and air fryer, obtaining 78 and 75%
Safety aspects, approval ratings, in contrast to the hotplate, which
received 89% disapproval across all households.
Ease of use and training aspects. Despite the general cooking speed, from the
data collected in the cooking diaries (ANNEX 1),
For the above qualitative indicators, the main cook it has been seen that several households cook
from each household would rate their experience large meals that require intense heating, such as
using the choices strongly agree, agree, disagree, preparing beans. The households were asked
strongly disagree and no opinion. The options were about the appliance’s suitability for large cooking
mapped to a numeric scale, centred at 0 (zero) to to understand the use of e-cooking appliances
eliminate numeric bias, whereby -40 = strongly for large meals. Based on this, the EPC obtained
disagree, -20 = agree, 20 = agree, and 40 = strongly an 83% approval rating, while the air fryer had an
agree, while no opinion was assigned no value. 88% approval rating for specific meals such as
Each question with the above-scaled responses marinated chicken. The hotplate received 93%
was then converted into a percentage. All opinions disapproval across the household, demonstrat-
were numerically scaled and assessed. ing slowness in cooking both normal and long
cook dishes. This slowness can be attributed to
The most liked feature across e-cooking losses due to the open heating element causing
appliances provided under study was their ability a low thermal coupling between the element
to regulate the heat, with the EPC and air fryer and the cooking pot.
receiving 67% and 75% approval, respectively,
45
Different e-cooking appliances handle the on the need for e-cook training and in which
issue of the cooking pot/cooking chamber areas such training should focus. All households
differently. The first class of appliances, spe- mentioned that training on electric appliance
cifically the hot plate, provides flexibility to use was not strictly required. A positive highlight
the household on the cooking pot used. The was the appreciation of the e-cookbook that
second class of appliances, such as EPC, are demonstrated how users could use the EPC to
engineered to be used with a specific volume cook local dishes. This permitted the transition to
of a cooking pot usually provided by the man- proceed without lengthy e-cook learning times
ufacturer. In the first class of appliance, the required by the households. As a result, the
unbounded nature of the cooking pot may be consultant was able to start collecting data from
a concern where the maximum volume of pot all households from the first day of the transition
owned is larger than what can be accommodat- phase.
ed on the cooking appliance – as a result, this
first class of appliances require proper market
research to be performed by the distributor/man- 4.2 User perceptions of
ufacturer to realize the benefits of such flexibil- electricity compared to
ity fully. EPCs and air-fryers though inflexible to
the cooking pots/chambers used, also require
baseline primary fuel
market research to ensure that the volume of the Participants were requested in the exit survey to
cooking pot provided is suitable for commonly provide opinions, comparing electricity to their main
cooked dishes. Assessment of cooking pots baseline fuel - LPG, across various criteria. Partic-
across the households yielded a 78% approval ipants responded positively to several aspects of
rating due to the large 8L volume of the ECOA electric cooking, including
pressure cooker. The 3.5 L air fryer obtained a
40% approval rating due to the 3.5 L volume. the speed of cooking in agreement with section
However, a 5.5L of the same brand exists. The 4.1 above,
hotplate received a poor 40% disapproval where ease of use,
it was not suitable for large cooking pots owned
by the household. As such, the flexibility of the the absence of smoke and
hotplate became a weakness due to its inability
the ability to cook irrespective of wind
to balance and accommodate non-flat and large
conditions which may affect other fuels with an
cooking pots.
open flame such as LPG or charcoal.
Safety is a key concern when working with
electric appliances, especially where children However, despite these benefits, participants
are involved. Of the three appliances provided mentioned grid unreliability, perceived higher cost
by the consultant, only the hotplate had an open and inefficiency of specific cooking appliances such
heating element. During the exit survey, partic- as the hotplate depleting the household’s electric
ipants provided opinions on the general safety energy balance as disadvantages to cooking with
of e-cook appliances. It was a common opinion electricity.
across participants of the safe nature of the EPC
and air fryer. The hotplate, however, received un- While comparing LPG with electricity, participants
favourable reviews across all households due singled out the reliability of LPG as its most sig-
to the exposed heating element causing safety nificant advantage. Participants also mentioned
concerns where there was a lack of adult su- that LPG is specifically faster than the electric
pervision. However, the EPC and air fryer were hotplate. However, the consultant notes that only 1
perceived as safe to use. (one) household said LPG is more affordable than
electricity. Additionally, the ability to cook multiple
Ease of use and training aspects. As part of dishes simultaneously in a multiple-burner LPG
the EPC accessories, an e-cookbook was stove is a desired feature compared to cooking
provided to all households. No e-cookbook was one meal in an EPC. Disadvantages to using LPG
supplied with the generic hotplate, while the include the constant need to purchase matchboxes
households that received the air fryer received that get misplaced regularly and the inability to
only a standard product manual. As part of the determine how much LPG remains while cooking.
exit survey, households were requested to opine
46
Across all participants, the LPG was used as an alter- average of two (2) dishes. Further, the exit survey
native energy source across all households when indicated that the households had an average of
electricity was unavailable. 50% of the households 3.4 cooking hobs or electrical cooking appliances
generally mentioned that they also perceived for parallel cooking.
electricity as slightly expensive compared to elec-
tricity. The other 50% of households believed it Other opinions include the ability to vent pressure
to be at par or cheaper than LPG. In contrast, automatically without requiring user interven-
one household believed that cooking with elec- tion throughout the venting process and having a
tricity was much more expensive than LPG. This suitable appliance to cook chapatti. Suggestions
indicates there is a general perception that cooking around hotplate design include the capability of
with electricity in Kenya is costly, while data show the hotplate to support both non-flat-bottom pots
otherwise. A household within the study group that and larger pots.
fully switches from cooking entirely with standard
fuels to electricity saves 0.11 USD per capita per
dish. This translates to 28.71 USD saved monthly 4.4 Addressing barriers to
in a household in Nairobi with 2.9 (KNBS, 2019) cooking entirely with electricity
household members, that cooks three meals daily.
Therefore, a 100% electric cooking transition would Table 29 shows the most significant barrier based
save 7.5% of a low-income household’s monthly on analysis of the reasons for not cooking a dish
spending at its maximum monthly spending of 382 using electricity using the cooking diaries. Too few
USD (KNBS, 2022) at the exchange rates provided suitable electric appliances, such as EPCs, exist
in this publication. This calls for more consumer during a cooking event. An intervention providing
awareness of the financial benefits of transitioning multiple efficient and suitable electric devices
to cooking entirely with electricity. would eliminate barriers 1 and 3 (hotplate is slow
and too much to cook), accounting for 39.4% of the
total barriers if the EPC has 100% suitability for all
4.3 User Perceptions and local dishes. Resolving electricity reliability would
Suggestions on e-cook further resolve barrier number two on electricity
unreliability - 22.3% of the total barriers. Control of
appliance features human factors such as forgetfulness, unwillingness
Participants were requested in the end-line survey to try new cooking devices and better planning of
to provide information on their general perception the household energy balance in prepaid arrange-
of the usability of the electric appliances. Most of ments through behavioural change programmes
the comments centred around the cooking pots and training would further resolve barriers 4, 5, 6, 7,
with the suggestion that a single EPC pot was in- 10 and 11 (listed in Table 29 above), yielding a 32.8
sufficient and households required multiple EPC decrease in the obstacles faced. Overall, the above
pots to perform sequential cooking. This further measures would ideally lead to a 94.5% decrease
builds on discussions noted in the earlier sections in barriers.
on heavier meals, such as supper requiring an
47
05
Conclusions and
Recommendations
5.1 Conclusions
Despite the training provided in the transition electricity. After the transition, each household
phase, the local cuisine-friendly interface of the would incur a monetary cost of 0.02 US$ per
EPC that contains ugali, among other foods, and person per dish, compared to 0.13 US$ per
an e-cookbook guide provided per household person per dish when cooking with LPG. These
– households did not cook entirely with electric- yearly cost savings can be realised sooner by
ity. More significant effort is required to demon- reducing the upfront cost of efficient e-cook
strate how dishes such as ugali can be cooked appliances by applying for carbon credits at a
using electricity. However, despite the difficulty project level under the Gold Standard14 and the
in shifting dishes to electric cooking, electric Verified Carbon Units (VCUs) programs.
cooking was noted to yield more complex meals.
Meals comprising three (3), four (4) and five (5) There are challenges to cooking with electric-
dishes increased in popularity while single-dish ity specific to a select type of technology. A
meals declined, which may yield better-bal- hotplate cook slowly has safety concerns as
anced meals. the heating element is exposed and cannot
accommodate round cooking pots. For EPCs,
The results showed that, on average, a household the need for specific cooking pots introduces
within the study group transitioning completely the need to keep emptying the first dish into
from baseline fuels to a 100% electric cooking a separate bowl and cleaning the pot before
target in the transition scenario with EPCs, cooking the second. This creates time loss and
hotplates and air fryers would reduce its is more evident if a household cooks more than
energy use from 3.68 MJ/person to 0.41MJ/ one dish. All three types of electric cooking
person per dish. This is in general agreement appliances tested in this study are limited in
with a 2019 study13 in Kenya that compared tradi- the quantities one can cook at a go, lacking
tional cooking practices to electric cooking that multiple cooking burners that simultaneously
found a median energy per person per heating allow several dishes and are incompatible with
event of 0.63 MJ and 0.83 MJ for the EPC and all local dishes. Other challenges with cooking
electric hotplate, respectively. When scaled entirely with electricity are unreliable electric-
to the average number of people catered per ity supply, shared electricity meters in rented
dish – 4, this data results in total energy con- houses (typical in low-income areas), and user
sumption of 2.52MJ and 3.32MJ, respectively. preference for prevailing cooking solutions due
The reduction in electrical energy noted may to convinces accompanying such solutions. The
likely be attributed to the minute-by-minute data inability to cook multiple dishes at a go can be
sampling used in the project yielding a clearer sorted by stacking electric appliances. Still, the
picture of appliance use and the households upfront cost of electric cooking appliances is a
utilising the e-cookbook provided to operate the bottleneck that must be addressed mainly for
EPC effectively, delivering less energy wasted low-income households. There is also a general
while testing. perception that cooking with electricity in Kenya
is costlier than other baseline cooking solutions,
Additionally, the study found that households although data collected for this study does not
would save 0.11 US$/person per dish, on support this belief.
average, after transiting to cooking entirely with
13. Leary, J., Scott, N., Numi, A., Chepkurui, K., Hanlin, R., Chepkemoi, M., Batchelor S., Leach M., Brown, E. (2019). eCook Kenya Cooking Diaries –
September 2019 Working Paper.
14. MECS (2022): New MECS & ClimateCare methodology for metered cooking devices approved by Gold Standard.
48
Participants applauded the availability of an for households to learn how to operate these
e-cooking book tailored towards local dishes electric cooking devices. This is more so
as opposed to a generalised manual on how to for appliances that deviate from the ordinary
operate the electric appliances. For example, an operation of standard cooking stoves. For
e-cook book is part of the products a household instance, procedures for using an air fryer and
receives after purchasing an EPC. It contains a EPCs differ from those required to use an LPG
recipe for dishes that can be prepared on the stove. Some manufacturers have identified this
EPC, instructions on how to set the EPC ready information gap and are providing e-cook books
to cook a dish and, finally, how to cook it using as accessories for EPCs (e.g. EPCs from Burn
the EPC. This reduced the time to learn how to Manufacturing). It is recommended that other
cook using an EPC and put the participants at manufacturers emulate this example.
ease, as it was not a try-and-error situation.
Research and development Most of the barriers
With 3 - 4 parallel-run electric cooking to cooking entirely with electricity, with the
appliances (The ability to cook more than one constant cost of appliances, is the limitation of
dish at a time) required per household, the electric cooking appliances. For instance, in-
availability of enough electric appliances to compatibility with local dishes, inability to cook
meet cooking requirements was not possible multiple meals at a go with one electric cooking
in the study, and households used standard device, and inability to accommodate large
fuels in parallel with electric appliances. The volumes of dishes (e.g. boiling large quantities
study found that electric appliances, including of cereals to last a week), among others. Some
EPCs, are not yet suitable for performing large- appliances still have issues with the time taken
batch cereal pre-cook activities, as noted in the to cook a meal, safety concerns and the inability
transition phase. As a result, households still to accommodate round cooking pots or require
use charcoal in such cases. There is a need to a specific cooking pot. Appliances that require
extend the scope of this study to investigate a specific cooking pot Need more research
these findings further. on improving these aspects of these electric
cooking appliances, which are critical determi-
nants for a household to entirely transition to
5.2 Recommendations cooking with electricity.
Consumer Education As discussed in this Expand the scope of the study One of the
report, cooking with electricity is constantly most significant barriers to transitioning to
evolving. We have more varieties of electric cooking entirely with electricity is the inability
cooking appliances in the market that are to cook meals simultaneously, the incompatibil-
more energy efficient, have enhanced safety ity of the cooking appliances with local dishes,
features, cook faster, and can cook various local the need for multiple cooking pots for cases
dishes. There is also data on time savings per where a specific pot is required and some in-
dish, energy and cost savings when cooking efficient appliances cooking appliances. It is
entirely with electricity. This information may not noted that dishes that cannot be cooked in one
be readily available to the end-users, as noted electric appliance can be cooked in another.
that there is a general perception that cooking For instance, chapattis cannot be cooked on an
with electricity is costlier than baseline cooking EPC but on a hotplate. However, households
solutions. Therefore, there is a need to conduct rated the hotplate unfavourably as a suitable
strategic consumer education that demon- cooking device. Current technologies indicate
strates how to use electric cooking appliances the suitability of an induction cooker as a direct
and discusses the benefits of transitioning to hotplate replacement. An induction stove avoids
cooking entirely with electricity. the open heating element by wirelessly trans-
ferring heat energy through electrical induction
Develop e-cook books for unique electric across metals. Due to the induction effect only
cooking appliances. In addition to consumer affecting metals, induction cookers are cool to
education, having an e-cook book that gives the touch even when on. However, a downside
a step-by-step guide on how to cook with to their use is the inability to use regular cooking
electric appliances improves the time required pots.
49
Nonetheless, providing multiple cooking pots bility of switching to cooking with electricity at
would address this challenge. The same applies 100%. There is also a need to expand the study
to EPCs that require specific cooking pots. A period to when the incentive of paid electrici-
combination of 2-3 electric cooking appliances, ty is withdrawn if households continue to cook
e.g. an induction cooker, an electric pressure with electricity or fall back to their baseline
cooker and an air fryer, would be more suitable solution. This would need to be accompanied by
for an electric transition if the cost does not affect consumer education discussed above. It would
demand. We propose to extend the study to also be interesting to probe the willingness
include these cooking appliances that tick most to pay for the upfront cost of electric cooking
of the desired aspects of a cooking solution appliances before and after using them for their
identified in this study and assess the possi- cooking needs.
50
ANNEX 1: COOKING DIARY FORM

ANNEXES
51
ANNEX 2: ENTRY SURVEY

Survey Data
1. Date
2. Name of supervisor
3. Name of enumerator
Information about the participant
4. Name
5. Gender
6. Age
7. Phone number:
8. What is the highest level of school you have attended?
9. Occupation
Information about the household
10. Group
11. Location
12. Type of area
13. How many people live in the household?
14. How many people cook in your household?
Cook details
15. Name
16. Relationship to head of household
17. What proportion of cooking do they do? (%)
18. When do they cook?
Information about the dwelling
19. How many rooms in the household?
20. Type of dwelling
21. Construction: Walls
Specify
22. Construction: Roof

Specify
23. Construction: Floor

Specify
24. Where is the kitchen?

25. Where do you cook?
26. What cooking appliances do you have in your house? Check all that apply.
Please specify:
27. Number of appliances selected: ${appliance_count}
52
Appliance
28. For the ${which_appliance}:
29. Take a photo
30. Brand or local name?
31. How many?
32. How often is it used?
33. What do you usually use it for?
34. How many hotplates/burners does it have?
35. Diameter of hotplates/burners (cm)?
36. Power in Watts
37. What fuels are you using for cooking now?
38. Do you buy or collect firewood?
Firewood
Buying firewood
39. How often do you buy firewood?
40. What quantity do you buy?
41. How much does that quantity cost? (KES)
Collecting firewood
42. How often do you collect firewood?
43. Where do you go to collect firewood?
44. How long does each trip to collect take?
45. How difficult is it to collect firewood? (1-Easy, 5 is Hard)
46. What was the cost of the firewood you purchased last?
Buying charcoal
47. How often do you buy charcoal?
49. How much does that quantity cost? (KES)
50. How hard is it to access charcoal?
51. What was the cost of the charcoal you purchased last?
Buying gas
52. When did you start using the gas cooker stove?
53. How much did you pay for the gas cooker stove?
54. What size gas tank do you use?
55. How often do you refill the gas tank?
56. How much does it cost to refill the gas tank?
57. How hard is it to refill the gas tank? (1-Easy, 5-Hard)
58. What was the cost of the gas refill you purchased last?
Buying kerosene
59. How often do you buy kerosene?
60. How much does that quantity cost you?
53
62. How hard is it to access kerosene? (1-Easy, 5-Hard)

63. What was the cost of the kerosene you purchased last?
Buying ethanol
64. How often do you buy ethanol?
65. How much does that quantity cost you?
67. How hard is it to access ethanol? (1-Easy, 5-Hard)
68. What was the cost of ethanol you purchased last?
Electricity
69. How long have you been cooking with electricity?
70. What do you spend on cooking with electricity per month?
71. What was the cost of the most recent tokens you purchased?
Perceived difficulty of using fuels
72. How difficult is it to cook with firewood? (1-Easy, 5-Hard)
73. How difficult is it to cook with charcoal? (1-Easy, 5-Hard)
74. How difficult is it to cook with LPG? (1-Easy, 5-Hard)
75. How difficult is it to cook with kerosene? (1-Easy, 5-Hard)
76. How difficult is it to cook with ethanol? (1-Easy, 5-Hard)
77. How difficult is it to cook with electricity? (1-Easy, 5-Hard)
54
ANNEX 3: EXIT SURVEY

Household Identification Number
Enumerator’s name
Respondent’s name
Enter a date and time
Location of household
Pre-survey task for enumerators
• Take the time to show gratitude to the participants for participating in this study for the last 2 months.
• Remember this is the last interview you have with the participant, so take your time as there are also a lot
of open-ended questions.
• Inform your participant: ‘If you remember any information about any of the questions as we go along,
please let me know so that I can note them down’.
1. Which electric cooking appliance(s) did you receive/purchase as part of this study?
Please specify:
2. Who in your family uses the electric cooking appliance?

Changes to cooking
3. Did you notice any changes in what food you cooked or the way you cooked it since you received your
electric cooking device? If YES, what were these changes? Why did you make these changes?
4. Have you learned any new tips/techniques/dishes since you started using these appliances? If YES,
what did you learn and how did you learn them?
5. How has the COVID-19 situation affected your cooking? Please explain.
6. Did responsibilities for preparing and cooking food change at all after the new appliance was
introduced? If YES, what were these changes? Why do you think these changes have happened?
Appliance review
7. How easy was it to learn to cook on the electric cooking appliance(s) you received?
8. Do people need to receive training on how to use an electric cooking appliance, or could they learn by
themselves? If training is needed, what should the training focus on?
9. What advice would you give others on how to use the electric cooking appliance(s) you received?
10. Are there any dishes from your usual menu (i.e. what you cook each week) that you could not cook
using the appliance?
11. Is there anything you would you change about this appliance? If YES, what would you change?
Experiences of cooking
12. What were your expectations of cooking with electricity before the study began?
13. Has this experience changed your mind about cooking with electricity? How?
14. What do you like about cooking with electricity?
a. What did you like the most?
15. What do you dislike about cooking with electricity?

16. Experience of using other cooking fuels.
LPG
Do you use this fuel for cooking?
What do you like about cooking with this fuel?
What do you dislike about cooking with this fuel?
55
Kerosene
Charcoal
Biogas
Firewood
Dung
17. Do you think cooking with electricity is cheaper or more expensive than cooking with the fuels you
normally use?
18. Do you think electric cooking is affordable?
19. Do you use either wood or charcoal (or both)?
Are there any problems caused by smoke?
What are these problems?
Are there any benefits of the smoke?
What are these benefits?
20. Do you think cooking with electricity is safer than with other fuels? Why? / Why not?
21. Did you have more free time to do other activities once you started cooking with electricity? If YES, what
other activities did you do during this time? Did you enjoy having this extra time? If NO, why?
22. Did you prefer cooking with electricity more than with other fuels? Why? / Why not?
Taste of food cooked using the appliance
23. Are there any foods/dishes that taste better when cooked with electricity? If YES, which foods/dishes
and why?
24. Are there any foods/dishes that taste worse when cooked with electricity? If YES, which foods/dishes?
And which fuel(s) would you prefer to cook these dishes with and why?
Fuel stacking
25. Have you continued to use other fuels since you started cooking with electricity?
a. Which fuels? And why?
b. Do you use different cooking fuels at different times of the year? Why?
56
26. Would you ever cook using only electricity and no other fuels? Is this something you want? Why? / Why
not?
Willingness to pay
27. Would you buy this appliance if you saw one in a shop?
28. How much would you be prepared to pay for this appliance?
Now tell them the respondent the price of the appliance (e.g. EPC 100$, electric Cookstove 70$) and ask
the next question.
29. Now that you know the price, would you be willing to purchase this appliance at this price?
Recommending
30. Would you recommend others to use or buy this appliance(s)?
a. Why?
31. Has anyone from your neighbourhood inquired about the appliance?
32. Have you recommended this appliance to others already?
Electricity supply
33. Did you have any power outages after you received the appliance? How often were these outages?
a. How long did the power outages usually last?
b. What were the main reasons for power outages in your area?
34. Did the outages affect cooking or heating water with the electrical appliance? If YES, how did you
cook??
Safety, repair & maintenance
35. Have you had any safety issues or safety concerns cooking with electric cooking appliances? If YES,
with which appliance and what was the experience?
36. Have you had any technical or maintenance issues with the electric cooking appliance you received as
part of this study?
a. What was the problem?
b. Did any of the issues above cause the appliance to fail at any point? If YES, which issue(s)?
37. What would you do if your electric cooking appliance(s) broke down?
38. If you had to replace your electric cooking appliance, would you replace it? If YES, what would you
replace it with?
Missing data
We have tried our best to learn as much as we can about how you cook, but we appreciate that the tools
we are using are limited. Please help us to understand what we may have missed.
39. Are there any meals that were cooked since the beginning of the study that were not recorded on the
forms you have given to us? If so, what were they?
40. Is there anything else that you think is important about the way you cook that we have not yet captured?
Study feedback
41. Has this cooking diary study been different to what you expected? How?
42 Did you find the training you received on using electric cooking appliance(s) helpful? Why / Why not?
43. Was the notepad useful? Why / Why not?
44. How did you find the enumerator visits and the interviews?
57
45. What could we do to improve the way we carry out the study?
46. This is the end of the survey. One final question: Will you continue to use your electric cooking
appliance(s)?
Finish the survey by showing gratitude once again for the participant taking part in the study.
OPTIONAL: Understanding the local way of cooking a key local dish
The example below was created for understanding the local way of cooking rice, but please edit and adapt
for the local cuisine: e.g. choose a different dish, change the answers.
1. What is the ideal quality of cooked rice?
Please provide any further details
2. Based on your answer to the previous question, what is the best way to cook rice to obtain the desired
quality?
3. What do you think about the belief that ‘cooking rice with firewood or charcoal tastes better than any
other fuel’?
58
59

An in Depth Exploration of Cooking Entirely With Electricity in Kenya

Uploaded by

Copyright:

Available Formats

An in Depth Exploration of Cooking Entirely With Electricity in Kenya

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

An in Depth Exploration of Cooking Entirely With Electricity in Kenya

Uploaded by

Copyright:

Available Formats

An In-Depth Exploration

of Cooking Entirely with

Document title: In-depth Exploration of Cooking Entirely with Electricity

© EED Advisory Limited 2023. All rights reserved

EED Advisory Limited

02. Approach and Methodology 4

03. Analysis and results 15

04. End-user perceptions and barriers to

05. Conclusions and Recommendations 48

Figure 1: Summary of Approach 4 Figure 22: Percentage usage of electricity 26

SDG7 Sustainable Development Goal number 7

The exchange rate at the time of reporting 1 US$ – KES 121

LPG market rate per kg – 1.84 US$

Electricity market rate per kWh – 0.176 US$

Charcoal market rate per kg – 0.743 US$

The calorific value of charcoal – 29.9 MJ/kg

The calorific value of LPG – 44.8 MJ/kg

The conversion efficiency of electricity – 3.6 MJ/kWh

Overview of changes noted in fuels used Overview of changes noted in cooking

In the transition phase, electricity was

witnessed reductions in cooking time of 47.7% Perceptions and barriers to cooking

The e-cookbook permitted the transition to Further suggestions to manufacturers –Most

Table 1: Summary of barriers

Table 2: Frequency of reasons for not cooking with electricity

1 Hotplate too slow 20

Conclusions and recommendations

i. assessment of the electric cooking devices,

Figure 1: Summary of Approach

2.1 Assessment of the available electric cooking devices

Visits to supermarket chains at specific 2.1.2 Determination of common dishes

Dried cow peas leaves

Water for tea

Dried veg with G.nuts

Free range chicken

2.1.3 Compatibility of available e-cook appliances with cooking processes

Table 4: E-cook appliances chosen to be provided. Source: EED

No. Type of electric stove Model Cost

2 Electric hot plate (single coiled burner) Generic KES 1,000 ($ 8)

3 Air fryer (3.5 L) VSYA35BBK KES 8,995 ($ 74)

2.2 Data collection

Access to electricity and quality of electric

Motivation: The consultant sought motivated

Location: Based on the consultant’s require-

b. Bio-ethanol measurements c. Electric cooking appliance measurements

Despite the efforts toward measuring ethanol use,

Figure 6: Data flow of minute-by-minute electrical measurement data. Source:EED

Consultant’s software on Linode

Sonoff Elite on Router Enumerator access of

Further, the consultant developed the inspection

2.3 Implementation of the surveys

Figure 9: The main data collection steps of the study

Entry survey Baseline Survey Transition Survey Exit Survey

Main activities Main activities Main activities Main activities

2.3.1 Entry survey

Figure 11: Baseline stoves owned across households. Source:EED

Table 6: Frequency of use of cooking devices within the baseline phase

# Household by surname Air fryer Charcoal Oven Gas M-gas Microwave

Table 7: Frequency of use of cooking devices within the transition phase

Table 8: Classifications of households by income group

# Household ID Household by surname Income group