Player is loading...
Workshop introduction by Jérôme Kämpf & Henning Müller + Keynote speech Pr. Dr. Guglielmina Mutani - Place-based energy models for sustainable cities
Pr. Dr. Guglielmina Mutani, Politecnico di Torino
Monday, Nov. 15, 2021 · 9 a.m. · 49m 26s
Urban areas and cities are increasingly interested in predicting spatial and temporal energy patterns of the built environment to foster their sustainable energy transition to a low-carbon future, also enhancing the quality of life. The main challenges are to guarantee energy security, energy equity and environmental sustainability, providing energy supply where it is needed. The optimization of energy consumption and production can be driven with a place-based evaluation of the energy efficiency potential of the building stock and the exploitation of the renewable energy sources locally available. Building energy-use models are used to represent and simulate building energy performance with a spatial distribution at different scales, taking into consideration all energy-related parameters: climate and local-climate conditions, urban context’s morphology, buildings’ typology and geometry, and occupant’s behaviour. The place-based approach using Geographic Information System (GIS) allows to collect data at different scale, geo-reference energy demand-supply data and manage energy-related variables from different databases. Among the current scientific energy models and techniques with a bottom-up strategy, it is possible to differentiate i) engineering models, ii) data-driven models (statistical or Artificial Intelligence, AI, based) and iii) hybrid models. The most used approach in literature is the engineering model, usually recognized as Urban Building Energy Models (UBEM), based on physical energy balance equations and a complete knowledge of input parameters. Statistical models involve regression techniques and are widely applied due to their simple structure and fast processing time. Data driven models based on artificial intelligence (Machine Learning, ML, techniques) allow to take into account complex patterns related to the real world (i.e., the stochastic nature of occupants’ behaviour or energy production with renewable energy sources), explaining mathematically the correlations between input parameters and building energy-uses. Integrating data-driven and physics-based methods (hybrid models) is increasingly of interest to researchers due to the possibility to explain the extracted knowledge of ML techniques as useful energy information and translating it into energy measures to be implemented. Energy modelling at urban scale is necessary to represents future scenarios of urban energy systems, evaluating the impact of climate change, also considering technical and social-economic constrains. Finding the suitable energy model with a place-based approach can ensure the best energy efficiency measures, guiding city planners and decision-makers to drive effective energy policies adapted to each specific territorial context.
Embed
Transcriptions
Note: this content has been automatically generated.
00:00:00
oh to this description
00:00:04
this year especially oriented words g. s. m. c.'s research scientists your research institute
00:00:14
it's cool organising professor for which is it's
00:00:19
good luck we'll sue sue also works of will come to you
00:00:24
look so we'll come works up to three months into rare form directly
00:00:30
puppies which which is really to bring people together and let them pork
00:00:33
bits were always have like focus topics for fruit and oh no he
00:00:38
would you be looking for the remote index due to to to be funny
00:00:44
really for the fruits you for the first to do
00:00:53
i think presentations with it
00:00:59
over these things you could do things like this so
00:01:08
cushions online book to consult to uh_huh it's questions from people blowing
00:01:14
so that's the presenters can cushions some cushions people or with this plan
00:01:24
i think the shuns lunch break over thirty it's also stay that's interesting discussions
00:01:35
it's it's presentations from
00:01:38
uh_huh crime as your research centre more to you it's it's
00:01:44
so you're pretty good if you just stay
00:01:50
uh_huh m. c. should get to some degree pricks
00:01:55
so i think we could start the exact so you're dealing
00:02:01
with the first presentation so it's my pleasure to a professor
00:02:08
i'm from c. d. review sheets or
00:02:14
um it's specialists it's it's shields
00:02:21
a full thank you first no thank you much with on these working the
00:02:30
okay with it
00:02:37
mm
00:02:43
uh_huh uh_huh uh_huh
00:02:51
okay yeah ah i uh
00:03:01
this one
00:03:08
okay
00:03:12
uh_huh okay okay okay okay
00:03:16
okay uh good morning everybody and we didn't mean uncanny
00:03:20
i'm mad and i'm working uh i'm researcher and i'm
00:03:24
working in the the department of energy it but it technically thirty now and today i'm going to present the
00:03:31
uh in um the the title uh i'm going to just the title laughter we're not there is no the the
00:03:38
is the the the crack the title using the program was well i'm going to talk about the plays bass that
00:03:44
um all they'll uh have for a sustainable cities and there s. in the communities uh
00:03:49
and after a little introduction i will uh uh illustrate the the most the significant the
00:03:55
uh models that we use the to represent the
00:03:58
energy consumption out of the buildings at human scale
00:04:02
and then i will describe some uh application uh uh of uh my current better search
00:04:10
and the emmy i i yeah i am i'm going to talk about the
00:04:16
plays bass that i mean energy model and a particularly i'm going to use the
00:04:20
a g. i. yes so too was or uh so too so we put a graphic information systems that
00:04:26
and the um usually i'm going to work or on a different label so
00:04:32
with information on the front scale about the uh the uh to localise the
00:04:36
um and then i'm going to talk about energy consumption model
00:04:41
uh so i'm going to describe the uh all the building with the the real track but the
00:04:46
sticks uh and the uh i'm going to use a basic valuables to a bill to my models uh
00:04:51
um then that uh i compare these uh energy comes on sean uh
00:04:57
uh we put the energy production uh sees them a big feasting ones are
00:05:02
and the uh the uh and my work finished with the um evaluation on
00:05:07
all the other no bubbling under sources of that uh i yeah but locally available
00:05:12
and uh i'm finally i use the uh uh the set the uh um yeah i
00:05:18
was i was uh all the might walk to the beach the images are making platform
00:05:24
uh will be a mid to uh reach a
00:05:26
energy sick you it a a weekly and environmental sustainability
00:05:32
and the my models uh the um the the
00:05:37
novelty of these certain that's not the novelty but the
00:05:40
yeah i'm gonna be a particular the particularly over the
00:05:44
this place based models is that the i use the
00:05:47
uh existing the database uh uh yeah generally use the by uh your mum planner so
00:05:54
uh and uh i'm available the information uh uh for example
00:06:00
of the beach that having model or did does or face model
00:06:03
or uh the uh technical maps uh uh for the c. d.'s or
00:06:08
uh we uh together with the of course at the climate the that
00:06:12
uh in the the energy consumption data for example from uh the sustainable energy
00:06:18
action plan or the new where sustainable energy imply map action plan
00:06:23
uh the same uh for the production or uh i use
00:06:26
of the information on a a a a billable online that
00:06:30
in the uh for the productivity uh or also uh i use the the
00:06:35
uh all the information about available energy source is that maybe label locally
00:06:42
of course at the mall the lead a ideal the uh our uh am depends on the
00:06:48
a main objective build up that uh search uh usually i
00:06:52
do a man fully or are really uh energy model in there
00:06:56
in the uh we've lisa plays bass the assessment the i can take
00:07:01
into account the all the constraints that that there are on the train
00:07:06
uh or on the c. d. a light at the technical but also environmental
00:07:10
economic and the uh the uh constrains the uh i'm influenced by the low
00:07:17
on the or regulation of the data as specifically in each uh c. b.
00:07:24
all of these um the ah use the to the beach but a
00:07:28
scenario uh at knock was scenario or a few there's future scenarios that
00:07:34
uh in the uh the day but take into account the all of the uh
00:07:40
uh the input data the modelling and then constrains uh on the
00:07:43
territory and saul uh yeah i can uh uh um evaluate the
00:07:48
the uh the real target select the a. c. can reach or
00:07:52
um yeah yeah but we the an energy targets
00:07:57
uh or i'm a mental targets are economic targets
00:08:02
yeah yeah first of all i'm going to talk about the the
00:08:05
uh type of approach uh that i can uh use uh a generally
00:08:10
yeah yeah i use a a top down approach or a bottom
00:08:13
up approach a and b. set type of of uh approaches uh uh
00:08:17
the pan the on the type of data mainly on the type data that that yeah but
00:08:22
usually if i yeah but uh energy they that uh if you're
00:08:26
been a or a nice upper scale uh i use a a
00:08:29
is of course a top down approach or and uh then uh
00:08:33
uh described being the the uh the the the built context but
00:08:38
yeah yeah i i'm uh even devise some simplified energy modelling the of the buildings are
00:08:44
take into account the all of the uh information that that yeah the egg cuban scale
00:08:49
and then i apply this model in this simplified
00:08:52
models i cuban scale with this i'm a correction uh
00:08:57
uh quite a few send the uh that take into account for example of the
00:09:01
local climate can be shown of the altitude or other energy
00:09:05
related the a valuable but i am the on the battery
00:09:10
and the bottom up approach uh uh uh use they they use the
00:09:15
um uh uh sunday today pay yeah but uh uh it the buildings gala
00:09:20
uh these a type of approach is more accurate that
00:09:23
i can uh describe abate the the energy use uh of
00:09:27
all beach building a in a c. d. in the uh of course uh i i can uh uh uh uh
00:09:35
make that are uh describe but the spectral distribution on all the characteristic
00:09:39
of each single a building yeah yeah and uh of the beat them adamant
00:09:44
uh the uh proceed u. i. e. afterwards is quite the
00:09:47
same my uh on uh depending on the type of the
00:09:51
a valuable that by yeah but uh a cuban scale
00:09:54
i i identify simplified energy models sell the buildings and
00:09:59
the no uh i applied the uh the small delay due bills kayla and in the end it with that
00:10:04
uh i use also some correction uh life you send that
00:10:08
to a specially vectors actually distribute the the energy consumption and then
00:10:14
uh i can uh matt but uh the uh
00:10:17
energy consumption uh uh with that the collected special distribution
00:10:23
about the model link yeah um yeah yeah
00:10:26
mm from um the the the literate or
00:10:31
uh the the energy mall there's the public buildings
00:10:34
can be divided in the process three even models are
00:10:37
uh that there may is the on uh usually are now
00:10:41
energy balance uh on uh the equation that describes the energy balance
00:10:45
uh of the uh building the all the a block of
00:10:48
buildings a or b. stick to our uh a c. p. um
00:10:54
yeah they are may is the on a or b. conservation of energy and the conservation of the mass
00:11:01
um afterwards uh um uh we have a data driven models that
00:11:06
uh yeah bet that are based on statistics are yeah
00:11:10
yeah and uh the y. which are wide the use the
00:11:14
use the and the uh a mutually they involve a redirection techniques uh uh
00:11:20
so are quite simple uh uh quite fast the uh and uh easy to use
00:11:26
uh and either a type of data that ivan mall there's a are the artificial intelligence i'm
00:11:32
the one that user to feature intelligence like a machine learning get that that even models that
00:11:38
um and that this type of models uh um uh
00:11:43
it can uh take into account also uh the complex uh
00:11:48
uh patterns uh all the real war the like for example
00:11:52
the the stochastic nature all the uh
00:11:56
the human behaviour or some uh uh um
00:12:01
uh um some feature all their noble and resources uh uh with uh for energy production
00:12:09
uh the we're working we had been in a red beast uh
00:12:13
yeah but the next future we're working on a hybrid models uh
00:12:18
which uh uh work uh on both the process a three
00:12:22
d. model but also dot at the drive and all those so
00:12:26
and now we are using statistical uh more statistical models about that
00:12:30
uh we yeah in and the next future we're going to use the also machine learning models a tool
00:12:36
uh um explain the make better uh some uh
00:12:40
uh valuables that uh i have not statistically a clear
00:12:46
and then that uh the calculate the the energy
00:12:50
consumption are compared with the uh energy consumption the yeah
00:12:55
in the us all uh we apply these uh the small there's a eh cubans k. let
00:13:01
uh uh with a place based assessment but like for example g. i. yes
00:13:07
this is a a the first the uh work uh that i'm going to present the uh yeah
00:13:13
uh it was made the for the research brought
00:13:16
brought that project citizen power uh in two thousand then
00:13:21
uh um we've been a central you are all well and the first the
00:13:26
um uh the conceptual schema uh can be described with
00:13:31
the a g.'s based on the the uh jennifer is not
00:13:36
the base or with all the information of the buildings uh yeah we're talking about that you would be in a in
00:13:42
um it where we have a a a big these to keeping me work
00:13:47
so we have a lot of data about the energy consumption would be the
00:13:52
uh some uh an object models uh all the the the most the
00:13:57
a typical uh buildings archetypes are also taking account the
00:14:03
uh data from later to like uh the research project popular
00:14:08
and then not with a uh with the basic types are applying
00:14:12
some quite a few sent the for example uh due to the
00:14:15
uh more or less a sardonic spongy shown a are more or less uh
00:14:20
um there s. m. sore face a green sore face it that way yeah but
00:14:25
uh in the front part of the c. d.'s are we apply the
00:14:28
small delay your been scalable describe but be in special distribution of energy consumption
00:14:35
another model that we have done a a a for a gram and
00:14:39
also that uh is the uh quite similar we start the with the
00:14:44
uh in anchorman doesn't argentina we stop the with the energy comes
00:14:48
on shown all but uh six of the party main towards uh
00:14:52
and in that graham and also up for a six years
00:14:57
uh and there we underlies uh the built environment that uh it didn't
00:15:02
define the p. p. carla and building the amazon nation building types or
00:15:07
and the uh we analysts also the strict selecting the beast
00:15:12
tinker with the uni form a a type of of buildings that
00:15:16
then that we uh i had to find the the uh energy intensity all
00:15:22
these uh the mean uh and the average i mean energy intensity of these uh
00:15:26
the piccolo archetypes are and we apply
00:15:29
these uh uh energy intensity a cuban scalar
00:15:33
eh using some quite few send the uh for example income and also we have a
00:15:38
a lot of difference in how did you the soul tool uh uh eh and ah the
00:15:45
our model uh it different altitudes that and then we compare
00:15:50
uh the energy consumption that we have uh with the uh
00:15:53
first uh a date that it the eh for the department those
00:16:00
uh they say uh where a top down a mall those uh
00:16:05
and here i uh we have it and i bought them out the model uh
00:16:10
um there that we apply the into lena uh with the in the project empowering there
00:16:16
uh and uh uh yeah we have a a a
00:16:19
lot of the data about the buildings of the area and
00:16:23
and the accurate data base so with the energy consumption of more than the two thousand buildings are
00:16:29
yeah we're just the the um databases are
00:16:33
uh i'm completely in the in the uh
00:16:36
data and uh uh uh removing the the
00:16:40
outlier with with the statistical that's a analysis so
00:16:46
uh then we um uh made the uh yeah we we didn't
00:16:50
divide the average uh uh energy consumption intensity of the buildings are
00:16:56
uh uh for example uh for a period of course track showing the but also all
00:17:01
uh depending on the compactness of the building with the soft face the volume of make sure
00:17:07
in the uh those type of uh uh um
00:17:11
all those so where a a great enough or to
00:17:14
uh evaluate also the eh em energy saving potential
00:17:18
that uh in the short to medium and long term
00:17:22
like for example yeah we can see that the for two arena uh which is uh
00:17:27
extol because see these uh uh with the a lot of course trains a technical economic carla
00:17:33
and environmental constrains uh uh because we have the and historical centre so we
00:17:39
uh it can all the uh do like traffic information all the the envelope
00:17:45
and we have uh some uh uh constrains the because we have a nice
00:17:50
to keeping make work so we cannot take minutes we cannot change uh the
00:17:54
uh um uh the boiler or uh of the buildings are
00:17:59
uh but uh as we can see a in another municipality around feuding know that
00:18:03
are much more much less a constraint so that's all we can group put together
00:18:09
a different the musicality to uh improve uh the
00:18:13
energy saving potential this is the main a concept that
00:18:18
yeah yeah there is another project the bottom up approach or do we have
00:18:22
done in a city motoring they say we choose it really to c. t.
00:18:26
and you have to we met with a lesser then a fibre thousands a or b. not with cancer
00:18:32
uh the principal is uh quite similar we did this
00:18:35
work uh yeah we've been the project to an italian uh
00:18:39
a project a zero energy building is mark cuban these to the
00:18:43
a call finance and the uh by the ministry of the
00:18:48
indication on a research in the unit in a university a in a in italy
00:18:53
and these uh the concept is quite similar we ab bit data
00:18:58
collection and integration database or a um we thirty i yes sir
00:19:04
uh we uh we can see that for example the pitiable because traction uh and the um
00:19:10
uh the surface to volume ratio of the residential buildings in setting up an is it
00:19:15
they know we need the and we uh we didn't divide
00:19:19
the the energy related valuables the more energy related valuables or
00:19:23
uh taking into account information available for each single buildings
00:19:27
saints it remotely may say in there we uh the the
00:19:31
the modelling and the validation and afterwards we applied that is a
00:19:36
more billing the egg cuban scale with the a. g. i. yes will
00:19:41
uh this uh after the after step one not a soul the data collection
00:19:47
uh we use adopted nine or which is a statistical software and to
00:19:51
evaluate the the the best um no um move people in every direction
00:19:57
uh um their valuables that the weather influencing the b.
00:20:00
energy consumption uh where the puerile construction the to the volume
00:20:04
the type of ground floor with come actual or a um
00:20:09
other type of the uh use uh uh for the ground floor
00:20:14
uh the occupancy rate to be a pampered or at least uh
00:20:17
is the amount the model of the type of proven the boiler power
00:20:22
uh we reach a equate percent of determination of zero point
00:20:26
eight before in the zero point eight p. three if we
00:20:29
do not consider the last two variables which uh we're not the available
00:20:33
for the we'll see about that as you can see there are some uh
00:20:37
uh these um uh this uh a good action is
00:20:41
not the would the for a very high volume buildings and
00:20:47
and we a control also the uh um significant p. v. t.
00:20:52
all the all these uh and multiple integration that with the p. b. you
00:20:57
for each scene don't learn valuable and the uh um which is a lower
00:21:03
than five percent uh and uh i know i have statistically you and here
00:21:08
are some of the others outside of the consign shown for each a single building
00:21:14
uh the same type of model that was applied the in it for the uh metropolitan city of around one in this
00:21:20
case of this work was made with the you a home
00:21:23
in collaboration with the recipe inside university of the a lot on
00:21:27
and we uh use the some mandate that uh about
00:21:31
the b. about the uh seven thousand of mailings uh
00:21:35
uh i'm from questioners or a given by a universal us up
00:21:39
into other on my uh it back the we use also uh the
00:21:44
east that that change was not the bayes and then we suppose they
00:21:47
can can maps to describe the the geometry of the built environment that
00:21:52
we we the peter some quite a few centre we evaluate the the more energy really valuable also for uh
00:21:59
the energy consumption model for space eating speak schooling that
00:22:03
the mystical what the production and the like the city
00:22:07
uh the uh move people in our direction that we obtain their uh yeah where
00:22:12
or what they would do with the and i uh quite percent of
00:22:16
the termination a higher than eighty percent there and the end with the significance
00:22:21
uh uh with that b. b. u. r. l. where than five percent
00:22:25
we applied then uh this uh uh uh you know a correlation on
00:22:30
uh on uh eh on uh the the uh on
00:22:34
the built environment at the municipal scale and we compare
00:22:38
the uh energy consumption uh with the big uh all the sustainable
00:22:43
energy action plan and then you wear sustainable energy and climate action plan
00:22:48
and then we did the evaluation and the the and g. i. yes uh
00:22:53
energy at last as the are there are some uh picture all the building that
00:22:58
the real cost action that's or face the volume racial the buildings uh and uh
00:23:02
the number of what we now we can separate boss volume that work you with the
00:23:06
for the different models and here are some examples of the early no interaction with a painter
00:23:13
b. c. is uh the one up for space eating the uh using the
00:23:17
the mm a buildup construction and this or face the volume of a show
00:23:22
yeah that out the p. b. use uh on the
00:23:25
these two yeah independent variables and the was significance uh
00:23:30
of the more people in every direction which are a little much lower than five percent
00:23:36
and they say is uh an example of the electricity and lee never get action analysis that
00:23:42
we we have made that uh we choosing from a set by the energy performance seen except
00:23:47
for the mystical what the production and the the energy performance thing that's a force be schooling
00:23:53
get any uh any and then we can see that being you for the both the uh um
00:24:00
a mario bozo in the uh the uh yeah analyses will be received well of
00:24:05
the body also be busy dual by uh these uh uh to uh independent valuables
00:24:13
uh the season and now uh we're going to talk about the
00:24:17
injury name all the sole physical base the model or processor based model
00:24:22
uh in this case that bases up top down model we start with
00:24:26
the the energy consumption not for the city of you enough for uh
00:24:31
a forty eight uh make sure it's or uh all but one kilometre are one kilometre
00:24:36
uh um and that we applied on engineering model
00:24:40
at a bayes the uh on a quasi state the
00:24:44
uh am amply uh an energy balance uh the
00:24:49
has a uh explain that by italian standards uh
00:24:54
uh i'm a a a eleven thousand a three hundred and um
00:25:01
uh as we can see we apply this type uh all the energy balance a a the beast that's
00:25:07
kayla not that to be a building scale using this
00:25:10
time of course variable like at the sky view factor
00:25:13
be a canyon effect there would be a wee
00:25:16
facial between buildings uh ended the main orientation of sweets
00:25:21
uh these uh a valuable were very important to explain uh uh
00:25:26
some data that we don't have the egg to human scale of course
00:25:30
b. and um the steps uh all these um mall be allowed
00:25:35
to create this model is quite similar to the one i explained before
00:25:40
uh first of all you have to collect to the that uh it
00:25:44
can be being a scaled a local building scale or in the sticks kayla
00:25:49
like uh the building a data the climate they that in the the new one morphology data
00:25:56
and we collect the uh these a database so with that g. i. g.
00:26:00
i. us to look to represent the um exactly the beach them adamant that
00:26:06
then that uh we uh apply uh we use of
00:26:10
this valuable also to create the uh the energy balance model
00:26:14
of course uh uh we use a a navajo enable with the scale
00:26:18
data so uh um uh we uh be the uh this valuable also
00:26:24
uh using also uh you been uh uh that you're one data
00:26:29
and we calculate the energy performance uh uh into it to make sure and we and
00:26:35
compare this may you with the major they that on the front uh of course uh
00:26:41
and sees on a man flea with the you know on the front seats on and that
00:26:46
uh if uh the two index so we're not comparable we calibrate again uh the of
00:26:53
course the valuable also we check uh the different it it flexes it to understand uh
00:26:58
the problem and the eh i've until uh the the the
00:27:03
the to the index were comparable and then uh we applied the
00:27:07
the model that you buy scale as an example here
00:27:11
we have the and describe you factor that we use the
00:27:15
uh they pick in we can a calculator cuban scalar with a g. i. yes tool
00:27:21
uh and of course the describe you factor in throwing so the solid gains are uh
00:27:27
back to it influence also the uh um
00:27:31
expert flax uh uh of the build environmentalist guy
00:27:35
and uh uh eh yeah beside the the skyway factor wheels also the average age the of
00:27:42
all beach single building and this main street orientation of uh at the block of billion scale
00:27:48
these are some results uh for two main issues uh uh uh uh
00:27:52
for space heating and the space eating until the makes the got water
00:27:56
of the friend that period of construction india are there are some results uh uh um
00:28:04
and eh it you can see uh the uh um and
00:28:08
mean absolute the uh um percentage arrow is quite low all
00:28:13
uh but higher uh yeah uh of course analogue can be absurd the
00:28:18
uh for uh some uh on p. pickle make shoes uh where
00:28:23
uh we have a low percentage of presidential build the connected to the you should
00:28:27
get the sticky the network or a high percentage all but during a any as well
00:28:33
a or a high may use all of i'll be door which are a and p. p. call for this uh type uh all the uh
00:28:39
uh on your one context the and
00:28:42
a high a building coverage racial goals of
00:28:46
the sky view factor and low they use all the relative a build the buildings
00:28:54
uh yeah uh the engine gonna mow the lawn which are we are working the uh uh now
00:29:00
uh is an hourly model and may is the uh
00:29:04
on a an energy balance uh uh all the three
00:29:08
after monday night make system that we have the on the built environment or
00:29:13
and the specific early but all park uh uh am
00:29:16
but all the transparent envelope the in the building there
00:29:20
uh like uh in padding on the structure for mutual object the people are and they are and
00:29:27
they would be so three a cease then uh
00:29:30
uh influence so wanting chad there and saul uh
00:29:34
uh uh it was build the any tentative the prophecy newark to solve the these uh energy
00:29:39
balance on that that these two sees them about quite simple with an exile a spreadsheet that
00:29:45
and be a input data are quite the same as uh the with the g. i. yes not the mesa
00:29:52
and uh we uh um a we did this by answer for each single
00:29:57
buildings take into account the all the
00:30:00
um uh all the um a specific
00:30:04
charred bodies to call each single building and we applied about the whip out
00:30:08
applied to this energy balance that it the uh um a block of building scale
00:30:15
and these are at that last those uh that we uh are
00:30:19
nice though we the information on on uh on the hourly uh
00:30:24
energy demand for spacey cleaning green but the uh as you can see we have the
00:30:29
all the the uh this is a game yeah and an example of a typical winter
00:30:36
uh they uh but we can see that we have uh all the uh eat flax seeds uh
00:30:41
for uh um uh the intended against or be a ventilation
00:30:47
it gains a and b. uh it losses by the uh
00:30:52
uh the uh am a transparent envelope or and the the op okay but up
00:31:00
in so we can understand uh uh if there are some arrows are
00:31:04
aware of the non uh whether that is a little look what i'm saying
00:31:08
uh the uh yeah a but i just sees another happy battle
00:31:12
with miss a model was uh uh about the fourteen percent that
00:31:17
uh with the uh back that it probably uh first the class there's a in we
00:31:23
too we have a more this is i make sure that it's not all but make sure
00:31:27
we from our buildings uh and uh i'm more uh and uh yeah more and more data
00:31:35
uh the aim of this uh work was to uh
00:31:37
the mm prepare are now is uh the project that
00:31:42
a a ditches you're making platform likely swam in week two yeah but
00:31:47
uh we can compare the uh chuck berry stick all baby stricter like this one
00:31:53
and then that to evaluated the energy targets of the environmental power gets that
00:31:59
uh the um but also that the remote conflict can be shown
00:32:03
that uh changing the some variables like for example the building being city
00:32:08
uh the uh the use of the rule for uh to produce energy
00:32:12
with for the what like the or sort of that not collectors that
00:32:16
uh um uh to uh use of the green roof uh with uh a green roof technologies are
00:32:22
in the uh who had compare uh uh the uh demo compo can be
00:32:28
shown a week then without the your one uh a green sort faces uh
00:32:34
and the same uh an hourly energy model was uh also validated the for the
00:32:40
a cooling the speech schooling the consumption uh in the here there is an example of a
00:32:47
quite a a clear example of the uh uh
00:32:51
for example or of uh how we and calculate
00:32:54
the the uh the shadows or faces a of the building name but opposite it the remote
00:33:00
and the block of building scalar take into account the for each out where the sun hate that
00:33:06
but also the i hate to we put a shot of your when cain young
00:33:12
and uh we of course we work all uh with the
00:33:15
same uh uh three patent time money not make six them sir
00:33:19
uh and uh we compare otherwise i'll sort of aware a g. i.
00:33:23
yes a model with the city semen and uh with uh i saw
00:33:29
uh i'm stand down the uh if if the two thousand all just a sixteen
00:33:36
we calibrate the model that will this k. let you know that our son uh uh keep because all the days
00:33:42
are a bit during the uh even if the you know it's not so all the as an italian c. t.
00:33:49
in the yeah yeah we uh then the air ones uh the the model was calibrated there
00:33:54
yeah we applied the tool of this model to p. p.
00:33:57
colour um block of building so that we had been to worrying
00:34:02
and always we see that that we can see that the model awards are quite whaler button all the
00:34:07
for uh uh the block of dealing with ours and which is that a
00:34:12
particular uh yeah not very used it in a in time and see these uh
00:34:19
and now we are working to improve uh the uh it flow rate
00:34:22
maybe inflation because uh in uh the previews uh uh example we use the
00:34:28
yeah yeah and and then i and have changes but ah why are
00:34:32
a constant the uh with a zero point three or zero point five that
00:34:37
and um depending on that that on the type of windows uh the be the the puerile construction
00:34:44
and so one a at the i had the aim of this work
00:34:47
was to then calculate the average amount plea uh had changes that i wear
00:34:54
uh also depending on the upper mobility all the yeah yeah of the windows are uh
00:35:00
considering that during the night the uh usually the permit let the upper maybe this law where
00:35:06
in that uh when uh the and the last scenarios when uh the
00:35:11
uh the and added changes per hour is to all probably we open the windows
00:35:19
oh yes here the last summer is not uh the first the uh baseball's first to work
00:35:24
was to uh uh was applied it to in a two two buildings are we different uh
00:35:30
uh orientation that uh considering their uh the main uh an
00:35:35
and when that action that and during the the winter a man for
00:35:41
uh and uh the uh i'm gonna be animal sitting here uh that
00:35:47
out of the results uh for a weaned exposed the building that this one
00:35:52
and then here we can see how the air changes but i work
00:35:56
changes the doing the the front mounted with many low we use that
00:36:00
uh and um uh in october and
00:36:04
uh in uh and december and january
00:36:08
and uh uh why uh these uh uh change but ours
00:36:12
uh changes by a forty one percent the uh it during the are the other man for
00:36:19
and we need the uh for different scenarios as i said before
00:36:23
and they hear out the main results are uh showing pay the uh
00:36:28
the the uh was the the changes uh all the air change
00:36:32
per hour as it can improve a white whale at the and yeah
00:36:38
energy consumption results uh a winning a c. p. like
00:36:41
you in with their they'll uh i win but also be
00:36:46
and finally they say is a a one album where last the work so
00:36:52
uh in which we compare uh in machine learning yeah i'm all they'll
00:36:57
and using light begun i didn't lose thing there with our dues basement
00:37:02
and angie adding model that uh and uh uh also we see this email
00:37:07
and uh we uh and just the uh where it ages based internally model with some
00:37:13
statistical uh improvement there and we apply these uh a comparison to the city of free border
00:37:21
and uh we need the am analysis so we
00:37:25
the modest base the uh to check uh uh it
00:37:29
yeah which are the battery about them that most the influence of the energy consumption and so
00:37:34
we i just the the name in the mall there's a take into account uh i uh
00:37:40
uh mainly these uh value bowser and we can say that the
00:37:44
the of changes but ah wire is quite important uh for uh
00:37:48
uh to analysis uh um uh they some although so
00:37:53
any uh our simon uh compares on all the three mall there's a
00:37:58
uh for uh uh some uh and uh they is a typical
00:38:01
day is a typical monthly day is uh in a framework that
00:38:07
uh the some some uh all the work uh and we have done
00:38:11
now with our model we apply the amount flee into any mall the lip
00:38:15
for example to check a which is the potential expansion all the indies to
00:38:19
keep in a worker a in the act while uh uh a scenario or
00:38:25
uh in a scenario we the uh in
00:38:29
a stand out the or advance the a vector fit into the main shown a
00:38:35
we've uh i'm in different than a worker in the
00:38:39
where we can and connect them are users of course
00:38:43
uh bases another application on on and off uh for
00:38:48
example uh um eh um we can uh yeah yeah
00:38:52
maybe using on the constraint that we had one that that
00:38:54
that eight thirty a. and define a priority winter main show not
00:38:59
and basing for example uh um our reality uh on the air
00:39:04
quality and uh yeah that out dark at the remote comfort conditions are
00:39:10
and the and in this case for example we applied the uh yeah uh
00:39:15
where the air quality is quite low and um we apply the potential during rueful
00:39:22
uh of course uh eh uh on the flak
00:39:25
through for a while at a uh in a
00:39:30
when the who feasible eh yeah is the way less or not exposed to
00:39:33
the air we can produce a energy or uh decided to use the the
00:39:39
uh i always make plans uh uh sir face about the uh with them an installation
00:39:46
uh we compare the uh yeah uh we're doing also some work uh with the outdoor
00:39:51
per remote comport can be shown all the our solution that this is an example of a
00:39:57
uh some uh an assimilation we do with the
00:40:00
n. b. make the comparing their a. t. pickled uh
00:40:05
i'm a summer day eh eh we fly
00:40:10
in without the and the green room for
00:40:13
so we can see how the surface temperature changes it
00:40:18
the aim of this war case to uh is again the to find that a and
00:40:23
a fast school shown not to be the some a solution
00:40:27
a good solution to improve the a our your by environment that
00:40:32
and we compare the results um in a yeah we the mean grad and pampered are we to use uh
00:40:37
for the outdoor environment to the main uh um the
00:40:41
body about that influence them and comfort can be shown
00:40:45
and we compare uh the and the mean that idea and pampered or uh uh
00:40:49
yeah uh well uh using city seem uh n. b. may the end sorbet get
00:40:54
in the s. simplify the equation uh uh with the just based information
00:41:00
and as we can see uh that are quite a big differences between uh the
00:41:05
um the mall there's about the uh um you how
00:41:08
word displays the uh equations seems to walk white whale
00:41:16
how much time where you have come up and then okay and of course uh yeah
00:41:23
yeah and it'd be and how warm all those are made that to a check uh
00:41:28
uh the energy performance in ewing um well first sees only eating the encoding season
00:41:35
in the uh in the ear in in uh
00:41:38
the last uh eh easier so we are working also
00:41:42
uh to evaluate which is the the best um a human morphology
00:41:46
because the um um the morphology of the built environment that that the
00:41:51
uh okay and then and uh and uh and the
00:41:54
n. best the on energy performance so but also uh uh
00:42:00
the best the morphology to produce energy with us all our uh uh technologies that
00:42:07
and they say he's an example or or or uh how we walked the uh the
00:42:12
season solid resources on the the rule for the building so that we can eh calculate that
00:42:18
of course uh is you can see is usual before we can uh uh do uh
00:42:22
the mall there's a about an eh electrical or energy consign shown we can make sure
00:42:28
uh the two information to get the the best the solution
00:42:34
and to produce a energy taking into account the the uh energy consumption
00:42:40
in their check uh uh the site sufficiency alright energy independence all the
00:42:45
uh that that eight thirty and also grouping the uh some uh a user so together
00:42:52
uh for example in and to communities are and and these
00:42:56
uh is an example all the game a typical d. strictly with
00:43:01
the dollars then shekels time shown for the what i can to
00:43:04
production and the uh the energy production out of also other users
00:43:11
uh yeah bases i yeah yes i'm all the uh evaluation that we have
00:43:16
done that and in the in the bottom graph for uh is you can see
00:43:21
and we uh and tried one allies the uh the uh
00:43:25
mm energy production not not only in the uh room for
00:43:30
uh and uh uh with uh on
00:43:33
south exposition but also uh this the not
00:43:37
marked the all of the north but i'm a rueful nor the oriented that for
00:43:43
and the two uh improve or the circles i'm sure now wires or
00:43:48
and uh and so uh this was eh um they're
00:43:54
one of the results of this uh analysis also not only
00:43:57
south or into the room for but also a part of the north the side that because uh that is mentioned newsletter
00:44:03
i am the a mm eh high consumption during the last the part of
00:44:10
the uh um uh of the day uh is spatially uh in the afternoon
00:44:18
and this is another uh uh uh um
00:44:23
walk we had than that uh we check uh in uh
00:44:26
some uh and uh we a analysis on p. p. carla
00:44:31
a beach the morphology in a six a the sixteen q. arena
00:44:35
and then we compare the compaq may solve the building there
00:44:40
we we uh we they energy production on a a potential
00:44:45
eh using the room for so with that a shopper back shown a to come so i'm sure i'm
00:44:51
and as we can see and of course we have a higher
00:44:55
and production to consumption or make sure i with higher also face the volume or a
00:45:01
show but the with high also face the volume racial we yeah we we have also higher
00:45:07
and number two losses are by a transmission and so we have to make sure these two information together
00:45:14
am i a we're doing yeah yeah some workshop
00:45:17
uh on this um all those uh every year
00:45:21
uh into arenas all uh uh they are open at this so workshop are open to all
00:45:27
the students uh uh i would like to participate in that i can conclude that with some uh
00:45:33
a strength and weaknesses uh all the values
00:45:37
approaches or a top down and bottom out but
00:45:41
and the slang to and weaknesses uh all the and different models and that we soul
00:45:47
and um i i will leave these uh oh yes
00:45:52
i can i comment a little bit okay yeah in
00:45:56
of course so that top down approach to is a very
00:45:59
simple and we have already all the information to do all
00:46:03
this type of approach or the the the weaknesses it is
00:46:07
the the uh difficulties uh in a um in a and
00:46:12
um differently uh apply the or the
00:46:16
special lie or uh information on uh
00:46:19
uh on the on a battery and because uh eh we don't have a uh
00:46:25
uh the mall there's usually do not take into account the uh special body of the the
00:46:30
uh if not the with the uh um simplified a connection quite a few
00:46:35
send the bottom up approach it is uh and much more accurate i that
00:46:40
uh of course we need a lot of big that uh and uh um we can
00:46:45
use the uh uh energy saving the also
00:46:49
a a a potential analysis with lisa data
00:46:53
uh of course of the data collection is uh along
00:46:56
the very beefy cool to and uh sometimes not possible
00:47:01
it it does certainly is not online uh the the data out this uh for this approach
00:47:08
and about the the models uh i can say that the then gina linger models uh is uh
00:47:14
might break for a the the because of my probably my formation and
00:47:22
yeah it lisa approach uh it doesn't need a lot the
00:47:25
a lot of data and and uh and consent to improve
00:47:30
their currency uh eh it specifically uh uh with an energy
00:47:35
balance or uh and that can that and and then uh
00:47:39
changing uh somebody most i can also simulate futures future scenarios that
00:47:45
of course uh and uh it doesn't take into account the or it is difficult to
00:47:51
take into account some uh factors uh like
00:47:54
a human related the fact also like human behaviour
00:47:58
a while uh and statistical uh is the most the use the uh model
00:48:05
and because the is the user friendly is very simple and fast that it'll apply
00:48:12
and machine learning yeah uh and could be a it
00:48:16
would the uh yeah it could be a good the
00:48:20
eh i'm i'm all they let because uh uh
00:48:23
uh we for these type of uh yeah uh um
00:48:27
a model that we can take into account the also some uh
00:48:31
i aspect of the award the uh of course uh eh
00:48:35
we need a very big big database uh and so uh
00:48:40
and and the of course uh as a all the mall there's a
00:48:44
the range of application is very important to check i i
00:48:49
would like to conclude that the of course uh uh yeah
00:48:52
i didn't talk about all my collaborators uh uh without the uh bremen i could
00:48:58
not the manager all these models and all this data on a soul of course
00:49:05
yeah i would like to think uh uh is not only my war case a workable do but and the
00:49:12
oh and uh a bit but you can join a if you
00:49:16
want but with our workshop uh each year in polytechnic little you know
Conference Program
Workshop introduction by Jérôme Kämpf & Henning Müller + Keynote speech Pr. Dr. Guglielmina Mutani - Place-based energy models for sustainable cities
Pr. Dr. Guglielmina Mutani, Politecnico di Torino
Nov. 15, 2021 · 9 a.m.
333 views
Q&A (Keynote speech: Pr. Dr. Guglielmina Mutani)
Pr. Dr. Guglielmina Mutani, Politecnico di Torino
Nov. 15, 2021 · 9:49 a.m.
Roberto Boghetti - Simulating District Heating Networks’ hydraulics with residual physics neural networks
Roberto Boghetti, Idiap Research Institute
Nov. 15, 2021 · 10 a.m.
Q&A (Roberto Boghetti)
Roberto Boghetti, Idiap Research Institute
Nov. 15, 2021 · 10:13 a.m.
Amine Weibel - Machine learning forecast for photovoltaic power stations (PV) with spatial interpolation on the COSMO-7 model. A real ‘use-case’ on the Czech Republic
Amine Weibel, HES-SO
Nov. 15, 2021 · 10:16 a.m.
Q&A (Amine Weibel, HES-SO Valais-Wallis)
Amine Weibel, HES-SO
Nov. 15, 2021 · 10:27 a.m.
101 views
Dr. Giuseppe Peronato - HeatLearn: a convolutional neural network to assess the building energy consumption at regional scale from open land-use maps
Dr. Giuseppe Peronato, Idiap Research Institute
Nov. 15, 2021 · 10:30 a.m.
Q&A (Dr. Giuseppe Peronato)
Dr. Giuseppe Peronato, Idiap Research Institute
Nov. 15, 2021 · 10:43 a.m.
Pr. Dominique Genoud & Jérôme Treboux - Integrating machine learning and AI in practical energy applications : a review of these last 10 years activities at the Institute of Information Systems
Pr. Dominique Genoud & Jérôme Treboux, HES-SO Valais-Wallis
Nov. 15, 2021 · 10:46 a.m.
Q&A (Pr. Dominique Genoud & Jérôme Treboux)
Pr. Dominique Genoud & Jérôme Treboux, HES-SO Valais-Wallis
Nov. 15, 2021 · 10:59 a.m.
Dr. Marco Tognoli - Fuel-saving potential simulation and estimation when optimally predicting the supply set-point conditions of a District Heating Network with detailed system data-driven modelling
Dr. Marco Tognoli, Politecnico di Milano
Nov. 15, 2021 · 11:31 a.m.
Q&A (Dr. Marco Tognoli)
Dr. Marco Tognoli, Politecnico di Milano
Nov. 15, 2021 · 11:44 a.m.
Pr. David Wannier & Jean-Marie Allder - Research advance in Energy prediction for electrical vehicles and green mobility hotels, the results of the eVIP project
Pr. David Wannier & Jean-Marie Allder, HES-SO Valais-Wallis
Nov. 15, 2021 · 11:46 a.m.
Q&A (Pr. David Wannier & Jean-Marie Allder)
Pr. David Wannier & Jean-Marie Allder, HES-SO Valais-Wallis
Nov. 15, 2021 · 11:58 a.m.
Dr. Kavan Javanroodi - Extending the concept of energy hub to facilitate sector and spatial coupling
Dr. Kavan Javanroodi, Solar Energy and Building Physics Laboratory (LESO-PB) at EPFL
Nov. 15, 2021 · 11:59 a.m.
143 views
