4 Years of Energy Management with openHAB
This is a modal window.
Das Video konnte nicht geladen werden, da entweder ein Server- oder Netzwerkfehler auftrat oder das Format nicht unterstützt wird.
Formale Metadaten
| Titel |
| |
| Untertitel |
| |
| Serientitel | ||
| Anzahl der Teile | 542 | |
| Autor | ||
| Lizenz | CC-Namensnennung 2.0 Belgien: Sie dürfen das Werk bzw. den Inhalt zu jedem legalen Zweck nutzen, verändern und in unveränderter oder veränderter Form vervielfältigen, verbreiten und öffentlich zugänglich machen, sofern Sie den Namen des Autors/Rechteinhabers in der von ihm festgelegten Weise nennen. | |
| Identifikatoren | 10.5446/61444 (DOI) | |
| Herausgeber | ||
| Erscheinungsjahr | ||
| Sprache |
Inhaltliche Metadaten
| Fachgebiet | ||
| Genre | ||
| Abstract |
|
FOSDEM 2023334 / 542
2
5
10
14
15
16
22
24
27
29
31
36
43
48
56
63
74
78
83
87
89
95
96
99
104
106
107
117
119
121
122
125
126
128
130
132
134
135
136
141
143
146
148
152
155
157
159
161
165
166
168
170
173
176
180
181
185
191
194
196
197
198
199
206
207
209
210
211
212
216
219
220
227
228
229
231
232
233
236
250
252
256
258
260
263
264
267
271
273
275
276
278
282
286
292
293
298
299
300
302
312
316
321
322
324
339
341
342
343
344
351
352
354
355
356
357
359
369
370
372
373
376
378
379
380
382
383
387
390
394
395
401
405
406
410
411
413
415
416
421
426
430
437
438
440
441
443
444
445
446
448
449
450
451
458
464
468
472
475
476
479
481
493
494
498
499
502
509
513
516
517
520
522
524
525
531
534
535
537
538
541
00:00
DatenverwaltungProjektive EbenePlastikkarteEnergiedichteGebäude <Mathematik>DatenverwaltungTermSoftware
00:55
GamecontrollerWasserdampftafelPhysikalisches SystemSoftwareProjektive EbeneEinflussgrößeMultiplikationsoperatorComputeranimation
01:26
Gewicht <Ausgleichsrechnung>EinflussgrößeStatistikDezimalzahlDatentypEnergiedichteEnergiedichteTermInterface <Schaltung>MultiplikationsoperatorStrömungsrichtungZahlenbereichFrequenzBus <Informatik>SchlussregelBenutzeroberflächeFunktion <Mathematik>BinärcodeOpen SourceAntwortfunktionExistenzaussageLeistung <Physik>Protokoll <Datenverarbeitungssystem>Physikalisches SystemVisualisierungTreiber <Programm>KorrelationsfunktionGamecontrollerRichtungInternetworkingFokalpunktKontrast <Statistik>Spezielle unitäre GruppeBitGraphRhombus <Mathematik>MinimalgradLokales MinimumGraphfärbungSoundverarbeitungPhasenumwandlungProjektive EbeneVirtuelle MaschineFamilie <Mathematik>RechenbuchBrowserBenutzerschnittstellenverwaltungssystemARM <Computerarchitektur>ZählenEin-AusgabeBitrateStrömungsdrosselDesign by ContractExergieStellenringFigurierte ZahlOffene MengeMathematikMeterComputeranimation
06:28
EinflussgrößeNeunzehnCASE <Informatik>Physikalisches SystemMultiplikationsoperatorMeterAggregatzustandSystemprogrammDatensichtgerätVirtuelle MaschinePotenzielle EnergieEnergiedichteDatenverwaltungDifferenteHardwareOrdnung <Mathematik>MAPStellenringEndliche ModelltheorieLastStandardabweichungKonfiguration <Informatik>Spezielle unitäre GruppeSoundverarbeitungARM <Computerarchitektur>Leistung <Physik>PlastikkarteEreignishorizontTypentheorieQuelle <Physik>Element <Gruppentheorie>KurvenanpassungDatenfeldEinfache GenauigkeitService providerTeilbarkeitComputeranimation
11:31
Leistung <Physik>EnergiedichteCASE <Informatik>SchaltnetzMereologiePhysikalisches SystemPrototypingComputeranimation
12:03
Ideal <Mathematik>Leistung <Physik>BitProgrammfehlerSystemprogrammCASE <Informatik>MomentenproblemNetz <Graphische Darstellung>MeterMAPSchlussregelOrtsoperatorSpezielle unitäre GruppeVirtuelle MaschineOffene MengeMultiplikationsoperatorOffice-PaketFirmwareJukebox <Datentechnik>SoundverarbeitungEnergiedichteMultifunktionMathematikMittelwertMereologieBitrateAggregatzustandPhysikalische TheorieARM <Computerarchitektur>ComputerspielGeradeSchlüsselverwaltungLokales MinimumGemeinsamer SpeicherCOMDiagramm
19:40
Flussdiagramm
Transkript: Englisch(automatisch erzeugt)
00:05
So let us get started. It's great to see such a crowded room here. I hope you're not all here just for the next talk to grab a seat. So my name is Kai, I'm a software architect and a project lead and founder of the project OpenHub.
00:22
I'm not going to talk about OpenHub for smart home today here in that talk. If you're interested in that project come to see me at our booth directly here in that building at the entrance. But I'm going to talk about more or less my personal story as a consumer, as an end-user in terms of energy management. My experience is there over the years and
00:44
the story actually goes much further back than four years. So it all started already 15 years when we built a house and I more or less electrified everything possible in there. So starting from the lights, okay, not that many people use candles nowadays anymore, sure,
01:03
but also heating is all electric, warm water through a heat pump, our photovoltaic system on the roof. So everything was nicely connected to a KNX system, so controllable I could get all measurements, but what was missing at the time was to really have some software that really helps me to visualize things, to
01:24
control things and so on. And that's why I started the OpenHub project in year 2010 directly as an open source project with the intention to have a system that allows me to create overarching rules and overarching user interfaces over all things
01:42
I have at home that have somehow an API that I can somehow connect to and to have such a system in place. Special focus in contrast to all the commercial solutions out there was on local control. I said, well, it's my home. I have all the devices at home there. They should talk locally with each other.
02:02
I want to have all my data locally and I don't want to have any dependency on the internet for that system. By now OpenHub grew quite a community and we have more than 400 different so-called bindings, which are more or less drivers for certain radio protocols. Other systems, technologies to reach out and to really combine into one single system.
02:25
And so you can more or less get everything what is available at home into that solution. Now, in terms of energy, what I did is that I hooked up such electric meters with an S0 interface in my electrical cabinet, which simply provide impulses as an output.
02:47
And I hooked them up to a KNX binary input, which then simply provide those on the KNX bus and created simple rules in OpenHub that count the number of ticks for a certain period of time
03:00
to calculate the current power out of that. And as you can see here from the graph, I have several of these meters. So one for heating and blue here, which usually really turns on and off load. Simply the green that was now one day last week here. So winter and not that sunny as the photovoltaic power produced.
03:23
And in yellow, the household energy that we use for more or less all the rest at home. Now, having that in the browser as a visualization is quite nice, but how to engage my family members to actually also get a feeling about the consumed energy. Well, we put up a fairly simple device here, an energy light,
03:43
which is basically an IKEA lamp with a Philips Hue bulb inside and a fairly simple rule in OpenHub that you can see here, which simply says whenever our household power changes, then if it's not night, because then we want all lights off, calculate U value ranging from green over yellow to red
04:05
and simply post that as a new color to that light bulb. And interestingly, that is a device that somehow goes a bit to your unconsciousness over time. So we're passing that many times a day or in the house and you suddenly feel after a while
04:22
that something doesn't seem to be normal. I didn't turn the dishwasher on or the washing machine and still it's showing red. So let's think what I might have forgotten. So it really gives a sense of and a feeling about the energy usage at home, not just for me, but also for other family members, which is a nice effect on that one.
04:44
Now for monitoring heating energy, this is quite a nice visualization, which shows a calendar here. This one shows last December, where you might remember here in central Europe, we had a very cold phase in the middle and it was fairly mild over Christmas and towards the end.
05:02
And so the background color here on each single day shows the minimum temperature of that day, ranging from minus 10 degrees that we had at home to, I think on New Year's Eve, it was around 11 degrees minimum temperature that day. And the diamonds here show then the used energy for heating that day.
05:23
And you see a very nice correlation here between those two figures. So that this can be also used to see whether everything works nicely or if you should actually check if something's not right. For monitoring photovoltaic system,
05:40
if you're set up a bit more complex graph that uses influx DB and Grafana dashboard, which both nicely integrate with open hub as a system to really get the data out here. So you can see in blue, the elevation of the sun for that day. In red, the luminance
06:02
in south direction. In yellow, then the power of the photovoltaic system, the gray bars show when it was raining that day. And so you really have a very nice visualization and you can check that everything's working all right. And also very good correlation here between really
06:22
light intensity and the photovoltaic power. So whenever something's off here, you could create alarms on your Grafana dashboard to actually say, hey, check, check your system, please. Luckily, so far after 15 years with that system, everything was smooth and I never needed a single alarm on that. Another nice event happened in spring 2015 when we had a partial
06:46
solar eclipse at home. And it was on a bright sunny day without any clouds. And that really resulted in a very nice curve here. And interesting thing is that with a partial solar eclipse, when you look outside, you hardly notice it because it's not going dark. It's still
07:03
daylight. But here you see that the power of the sun really went down by a factor of three to four roughly. And it was almost as if it's dark. So it was quite a nice effect. So all the monitoring is nice and good. But in the end, when you're talking about energy
07:22
management, you really want to do some, well, load shifting, optimizing your consumption and all of those things. Now, unfortunately, at the time that our photovoltaic system went live, at that time, there was no incentive at all for the end customer to self-consume
07:43
that energy that is produced. But everything goes to the grid and it's paid there and that's it. So there is no benefit for me to actually shuffle around some loads and do things. So my only option was to say, well, OK, our utility should provide different price levels over the day and I can maybe shift things for that. And thinking 10 years back,
08:07
the standard example for shifting load was, hey, you can do your washing at night. That was what everybody came up with, more or less. And so I said, well, OK, sounds interesting.
08:21
Let's see, such a washing machine that was smart grid ready usually cost around 300 euros more than the same model without such a feature. OK, you could say, well, one time investment. Let's go for that. Fine. And at the time, also in Germany,
08:41
the utilities were legally obliged to offer you at least one smart tariff that had to have two different price levels, at least. So I said, OK, let's check that out. And my local utility said, OK, we have a field trial here. And in order to participate in that, you actually have to book our smart tariff, which was an additional 100 euros a year.
09:05
I have no clue why, because we already had a smart meter. So there was no hardware investment or anything involved in that. But they provided an API then, which said for the next day, for that hour of the day, it will cost you that much money. And the price difference between
09:21
high and low was exactly three cents per kilowatt hour. So I quickly checked, OK, washing machine, what does that mean actually as a yearly consumption? It's roughly 150 kilowatt hours that you assume here. So I did some quick arithmetic and came to the conclusion that,
09:42
hey, you can save four euro fifty a year by doing all your washing at night. And yeah, so that doesn't sound that much, but you might now argue, OK, you can also use your tumble dryer at night. You could maybe wash your dishes at night as well and maybe even move your
10:02
warm meals to the night when everybody else is asleep. But even then, you're not coming anywhere close to actually have any benefit from all of that. OK, so that wasn't too interesting for me, unfortunately. And somehow my local utility also noticed after a while, hey, that doesn't seem to be too attractive. Nobody really wants that. And actually,
10:23
they came by and told me that, hey, those smart meters that you have at home, they break so often and then they can't read the LC display anymore. And so they can't get the number out of the meter and they have no clue what to do about that. So they said, well,
10:41
in 2016, they ripped that out and replaced it by an old school Ferrari's meter and said, hey, that one is really lasting 10 years. We don't have to come by. Everything fine. So here you go. So that was it, more or less, with all my attempts being really in the front there,
11:02
doing energy management and trying to be cool with all the smart home stuff and automation here. And that stayed like that until more or less four years ago when we bought this nice little blue Tesla here, which had a huge battery. And I thought, OK,
11:20
so much battery to store energy. I have to do something with that now. As I said, photovoltaic system wasn't really helping me here because there was no incentive for self-consumption. So I had to put up a second photovoltaic system, this time on the garage roof. And in 2019 was now the case that for this one, giving power to the grid
11:41
hardly gave you any money. So you had a big incentive in using all that energy yourself and optimizing that really. And so, yeah, big parts of the household energy during the day is automatically covered then by such a photovoltaic system. But then with the combination of the car,
12:03
surely surplus charging becomes very attractive here to say that everything that exceeds what you need in the household should be used for charging your car. Quite luckily then for more or less the pandemic times was that, well,
12:21
everybody did home office. So did I. So the car was at home during the day when it was sunny. So that worked out really well. And this here shows now another open hub rule that simply says that whenever the photovoltaic system power changes or the household power changes,
12:43
then please check if the car is connected to wallbox and adjust the current that the wallbox is delivering to the car. And I have a Keba wallbox that accepts UDP packets here to control it down to a milli-amp granularity, which is really nice because you can steer it
13:04
very precisely here. You have to at least go with six amps though, which is more or less the minimal power to start charging of the car. But with that rule, I can do all of that. And on the next slide, you see more or less than the outcome on a very nice sunny day. So in blue,
13:26
you have here the overall power that goes to the grid or comes from the grid. And the idea is to really level that out on the zero line, ideally. So in the morning, when there was no sun, we had to draw power from the grid. Then the sun came up or we
13:46
gave some power to the grid until the car started to charge then up a certain level. And then you can see that it's really fairly flat at zero. So that works pretty well. Then came lunchtime when the household power consumption was a bit more bumpy going up and down. So it's a bit more
14:06
tricky to level that all out, but it worked also quite well. Then I think at the end, dishwasher went on, which used so much energy already that the charging had to stop completely. And it turned out that there was some bug in the car firmware that didn't resume the charging
14:25
afterwards anymore. So at that time, I had to manually then always go there and have to restart it. Luckily by now, this bug is fixed by Tesla. And yeah, so the rest of the day, the charging rate was a bit reduced and yeah, it works quite well. And overall,
14:47
you can see that on the next slide, that's the yield of the photovoltaic system over all of last year. And an average that was between 10 and 11 kilowatt hours per day.
15:01
And if you consider that half of that, so five kilowatt hours is then used for the charging that corresponds to roughly 10,000 kilometers a year of driving the car. Obviously a bit more in summertime and not that much in wintertime, but it corresponds to a saving of roughly two tons carbon dioxide, which is quite a nice effect
15:28
here. And yeah, that's my experience so far. I'm looking in the future to also integrate with other solutions like EVCC, for example, which specifically looks into
15:42
car wallbox monitoring and also going like open staff, which is sounded quite nice into looking into the future, predicting and getting more machine learning stuff in there, which might be a topic for next year then. And with that, I thank you very much for your
16:02
attention. Are there any questions? Please.
16:33
Okay. The question is, if I can imagine whether I more or less give control more to
16:41
the grid operator than controlling it myself. In theory, I can imagine that, but from all that I've seen out there is that that's still a far, far future that really the utilities would be in a position to really make use of that data. And a problem that I see is also how
17:08
do you actually make sure that the data is real, that I'm not just giving anything there for maybe benefiting in some way of a better tariff or whatever.
17:21
And sorry, it's measured by the meter. Okay. If that's all just the pure meter values, that's I think anyhow already possible with the smart meters that are installed. Not in my case now at the moment anymore.
17:58
For having the utility allowed to decide when to charge and discharge the car,
18:04
I still want to be in the position to say, well, I actually need it at that charging state at that moment and so on. If that can be fulfilled, that works there. Washing machine is also something that really goes into your own personal comfort a lot or if they decide when
18:20
to do it. And so it's all a bit tricky. I think it's better in the households to really decide what to do and give incentives to do the right stuff.
18:49
Yeah, no, from utility and the grid side, it's obviously very important to not see a single
19:07
household, but to see more or less a whole city, part of the city and so on, and to be able to control things there to more or less get a decent level. That's for sure, but I think it's helpful to provide incentives to the single people
19:23
by having an API to interact with and then that might work. Okay, I see my time's up. Thank you very much. If you want to discuss further, I'm at the booth.