the purpose of this is to roughly right determine the daily average power peaks per month as used in Sweden (and other places maybe including Beligium?, Netherlands?). I used the method defined by Ellevio in Sweden here:
https://www.ellevio.se/abonnemang/ny-prismodell-baserad-pa-effekt/
(previously I called this code by Ellevio but have renamed it to a general name EPPMA and did some serious refactoring (beware))
The hope is to ultimately use this for load shedding / balancing / planning by being able to set a maximum allowed energy use each hour (based on the usage peaks and prices) to keep the peaks and prices lower.
After a few days, the data collection is all now seemingly working pretty well.
graphically in apex charts it is looking like this for checking/debug (previously):
latest albeit without much data at the moment as I just cleared things in the update(s):
It is quite a few steps and I am using Apex to present it so far. It has recovered well with various hiccups and I have now (230521) checked this vs Ellevio in detail and it has been preforming well but it will not be 100% as the import data I have is not 100%, again the idea is to get roughly right.
if you want to try have a look, see below install and process to understand
help, advice, tips, improvements?
I am guessing as I am not so great at jinja/yaml that there are some improvements to be made here, some things I have identified 
some things I see you can find in the open issues in the TODO section of the header
give it a try! here is the install/config
install and code incl apex chart config
pre-requisites
import power is needed and as it comes in it is integrated to kWh… if you have it already in kWh in a good enough time period you should be able to just put it into step 3! Mine did not run so often so I went to import power.
NOTES
see notes in code incl process flow and TODOs (improvements?)
configuration/install instructions
1. (of 4) put in the line in your configuration.yaml:
homeassistant:
packages:
eppma_energy_peaks_per_month_average: !include packages/eppma_energy_peaks_per_ month_average.yaml
2. (of 4) put the package in your package directory “package” with the name described in yaml (you can change director but you need to change it the config.yaml above)
## ENERGY PEAKS PER MONTH AVERAGE (EPPMA) ((EFFEKTTOPP))
## NOTES
# 0. based on ellevio information here https://www.ellevio.se/abonnemang/ny-prismodell-baserad-pa-effekt/#h-exempel-elanvandning-och-effektavgift
# (many other electric companies use this depending on country - sweden and maybe belgium and the netherlands.
# variables : this allows you to
# a. adjust the number of peaks by changing the initial dictionary size in eppma_7 add a row of fake data or remove them
# b. adjust the time of day for reduction in peak values in eppma_3 now 22:00 to 06:00
# 1. naming convention : aeppma_*#* where # is the step in the process
# 2. time shifts : values for hourly sensors (eppma_4 and above) are +1hr (and in the case of sensor values for the day it is next day)
# until reset the following day. This is so the sensor can take into account the full hour values
# eg values from 11:00-12:00 eppma_4-5 will shown between 12:00-13:00
# and eppma_6 & 7 will give values on the day after, (for 6 it is even just the first day of the month)
# eppma_8 is the previous months value
# 3. BUG average of peaks : eppma_7 - NOTE! there is NO feature to chcking having duplicates! do not run multiple times it can overwrite already placed data over data that is pertinent.
## process flow
# 1. import power -> integrate it to kWh.. i have power in watts coming in this should give kWh
# 2. utility meter #1 - getting an hourly kWh usage
# 3. find the max of #2
# 4. every "hours: /1" take #3 and adjust the values to take into account off peak and save - this is the hourly average (again this is showing data on the hour following the data measured)
# 5. get the rolling maximum of #4 per day - this is at end of day the peak during the day
# 6. at 00:00 and 5s and after #5 runs return the maximum per day of #5 - this is the daily maximum (saved on next day unfortunately)
# 7. find the 3 peaks highest of #6 per month and average - 3 peaks per month average
# 8. once per month get last months value from 7 and save it during the current month
# ---- i have gotten this far -----
# 9. ?? multiply #7 by consumption to get actual cost impact (?)
## CHANGES keep previous and this version change list
# 250312 - initilization of 7, fixed value sort of 7 to get everything over 0 not 0.001
# 250320 - changes incl changed name from ellevio to eppma
# made this a package instead of random stuff
# correction on ellevio_ version removed average on 3 and made it max value instead
# fixed triggers instead of time
# included datetime data in eppma_7 as output to make it easier to read directly
# added 8 a monthly keeper of last months data
## TODO
# 250307a time shifts -see comments in the code but the data you see > eppma_3 is not applicable to the zone it is found, it is an hour or 1 day or even 1 month after.
# 250307b sensor names - maybe i should have named the sensors better to address time shifts (?). eg.. it is not the hourly average now it is the hourly average of the last hour .. but if time shifts are fixed then they are maybe ok? eg in Apex I plot them and can add an offset there and it is solved.
# 250320 - adjust data by 1hr? .. maybe actually output in eppma_7, correct time in 7?
# 250321 triggering - i am still not happy with the triggering and timings.
utility_meter:
## UTILITY METER ENERGY PEAKS PER MONTH AVERAGE (EPPMA)
eppma_2_energy_per_hour:
name: "EPPMA 2 Energy per hour"
source: sensor.eppma_1_import_power_to_energy_spent_integral
unique_id: a7a703a5-4894-4ed0-a6c6-56392c488475
cycle: hourly
template:
## TRIGGER SENSORS ENERGY PEAKS PER MONTH AVERAGE (EPPMA)
- trigger:
- trigger: time_pattern
hours: /1
minutes: 0
seconds: 0
# action:
sensor:
- name: "EPPMA 4 energy hourly value adjusted for reductions during non peak hours"
# name: eppma_4_energy_hourly_value_adjusted_for_reductions_during_non_peak_hours
# sensor.eppma_4_energy_hourly_value_adjusted_for_reductions_during_non_peak_hours
## value saved here is the previous hours value
unit_of_measurement: "kWh"
unique_id: f5e57a8d-ff03-4d65-bc90-8c4c044916b2
state: >-
{# ref https://www.ellevio.se/abonnemang/ny-prismodell-baserad-pa-effekt/
""
Så fungerar det
Effekten mäts som timmedeleffekt, vilket innebär genomsnittet av din totala elanvändning under en timme.
Effektavgiften baseras på snittet av de tre högsta effekttopparna, det vill säga de tre timmar med högst timmedeleffekt, fördelade på tre olika dygn under månaden.
Varje dygn mellan klockan 22 och 06 på morgonen är det dessutom billigare att använda elnätet. Då räknas bara halva effekttoppen.
""#}
{% set hour_test = now().strftime('%H') %}
{% if hour_test >= "22" or hour_test <= "06" %}
{{ (states.sensor.eppma_3_energy_per_hour_max.state | float / 2) | float }}
{% else %}
{{ (states.sensor.eppma_3_energy_per_hour_max.state | float / 1) | float }}
{% endif %}
- trigger:
- trigger: time
at: "00:00:05" #delay 5s to make sure the last hourly value from eppma_5 is available.
- trigger: state
entity_id:
- input_button.eppma_6_trigger
sensor:
- name: "EPPMA 6 energy peak per day"
# sensor.eppma_6_energy_peak_per_day
#entity_id: eppma_6_energy_peak_per_day
unit_of_measurement: "kWh"
unique_id: 69fc7eb1-12d6-4adf-b8de-04faafb7c3f3
state: >-
{{ states.sensor.eppma_5_energy_adjusted_daily_max_peak_value.state | float }}
attributes:
date_last_evaluated : >-
{{ now().strftime('%Y%m%d%H%M%S') | float }}
- trigger:
- trigger: state
entity_id:
- sensor.eppma_6_energy_peak_per_day
attribute: date_last_evaluated
- trigger: state
entity_id:
- input_button.eppma_7_trigger
action:
- variables:
results: >
{# check value each day and see if it is higher than any of the items in the array of highest values if so then replace the appropriate data point #}
{% set ev = states.sensor.eppma_6_energy_peak_per_day.state | float %}
{% set evts = as_timestamp(states.sensor.eppma_6_energy_peak_per_day.last_updated) %}
{% if state_attr('sensor.eppma_7_peaks_per_month_average_per_day', 'data_timestamp') == none or (now().strftime("%d") | int == 2 and now().strftime("%H") | int == 0 ) %}
{% set in_dict =
{ -0.0001: '946688400.0',
-0.001: '946688400.1',
-0.01: '946688400.2' } %}
{% elif state_attr('sensor.eppma_7_peaks_per_month_average_per_day', 'data_timestamp') != none %}
{% set in_dict = dict.from_keys(states.sensor.eppma_7_peaks_per_month_average_per_day.attributes.data_timestamp) %}
{% endif %}
{% set in_eid = in_dict.get(in_dict.keys() | sort()
| select('<=', ev | float(0))
| first, -1 ) %}
{% set ns = namespace( kvps=[], d={}, l=[] , ldt=[] ) %}
{%- for key, value in in_dict | dictsort() %}
{% if in_eid == in_dict[key] %}
{% set ns.kvps = ns.kvps + [(ev, evts)] %}
{% set ns.l = ns.l + [[ev, evts]] %}
{% set ns.ldt = ns.ldt + [[ev, as_datetime(evts).strftime('%b %d %Y @%H:%M - %Y%m%d') ]] %}
{% else %}
{% set ns.kvps = ns.kvps + [(key, value)] %}
{% set ns.l = ns.l + [[key, value]] %}
{% set ns.ldt = ns.ldt + [[key, as_datetime(value).strftime('%b %d %Y @%H:%M - %Y%m%d') ]] %}
{% endif %}
{% endfor %}
{% set av = dict.from_keys(ns.kvps | selectattr( 0, 'gt', 0.000 ) | list) %}
{{ av | average | float, ns.l | list , ns.ldt | list }}
{# results to send, like after day 1:
(0.4361
, [[0.4361, 1742561081.363402], [-0.001, '946688400.1'], [-0.0001, '946688400']]
, [[0.4361, 'Mar 21 2025 @12:44 - 20250321'], [-0.001, 'Jan 01 2000 @01:00 - 20000101'], [-0.0001, 'Jan 01 2000 @01:00 - 20000101']])
#}
result_average: '{{ results[0] }}'
result_timestamp: '{{ results[1] | list }}'
result_datetime: "{{ results[2] }}"
sensor:
- name: "EPPMA 7 peaks per month average per day"
# sensor.eppma_7_peaks_per_month_average_per_day
#entity_id: eppma_7_peaks_per_month_average_per_day
unique_id: 57ce4d4e-f8d9-412f-9d87-82b08e0bc462
state: '{{ result_average }}'
attributes:
data_timestamp: '{{ result_timestamp }}'
data_datetime: '{{ result_datetime }}'
- trigger:
- trigger: template
value_template: "{{ now().day == 1 and now().hour == 1 and now().minute == 1 }}" #on the first of the month after the data has been run through.
- trigger: state
entity_id:
- input_button.eppma_8_trigger
sensor:
- name: "EPPMA 8 peaks per month average last months values"
# sensor.eppma_8_peaks_per_month_average_last_months_values
#entity_id: eppma_8_peaks_per_month_average_last_months_values
unique_id: 4fdea9cc-d23f-48a6-beb9-f06aebd9cd3a
state: >-
{{ states.sensor.eppma_7_peaks_per_month_average_per_day.state }}
attributes:
data_timestamp: '{{ states.sensor.eppma_7_peaks_per_month_average_per_day.attributes.data_timestamp }}'
data_datetime: '{{ states.sensor.eppma_7_peaks_per_month_average_per_day.attributes.data_datetime }}'
sensor:
## SENSORS ENERGY PEAKS PER MONTH AVERAGE (EPPMA)
- platform: integration
# name: eppma_1_import_power_to_energy_spent_integral
name: "EPPMA 1 Import power to energy spent integral"
unique_id: fdb30b8d-edfe-4bdc-b72a-2fa790422344
source: sensor.import_power
unit_prefix: k
round: 4
max_sub_interval:
minutes: 5
- platform: statistics
name: "EPPMA 3 energy per hour max"
# name: eppma_3_energy_per_hour_max
## max at end of period is max for previous hour
unique_id: e4be715b-cfb4-42e7-bc1f-e20248e44eda
entity_id: sensor.eppma_2_energy_per_hour
state_characteristic: value_max
max_age:
hours: 1
precision: 4
- platform: statistics
name: "EPPMA 5 energy adjusted daily max peak value"
# name:eppma_5_energy_adjusted_daily_max_peak_value
## max at end of period is max for previous day
unique_id: 168f3991-8e38-4d55-8b57-4e519cb8f4e1
entity_id: sensor.eppma_4_energy_hourly_value_adjusted_for_reductions_during_non_peak_hours
state_characteristic: value_max
max_age:
hours: 24
minutes: 0
precision: 4
input_button:
eppma_6_trigger:
name: EPPMA 6 trigger
eppma_7_trigger:
name: EPPMA 7 trigger
eppma_8_trigger:
name: EPPMA 8 trigger
#manual trigger - TODO -- DELETE after testing (?)
Apex code
type: custom:apexcharts-card
update_interval: 2.5min
header:
show: true
title: EPPMA energy peak /mth ave - dashboard
show_states: true
colorize_states: true
experimental:
brush: true
brush:
selection_span: 1d
apex_config:
chart:
height: 300px
graph_span: 70d
span:
start: day
offset: "-69d"
yaxis:
- id: first
decimals: 2
- id: second
opposite: true
decimals: 1
series:
- entity: sensor.eppma_1_import_power_to_energy_spent_integral
name: 1. importPower integrated to energy
yaxis_id: second
stroke_width: 1
extend_to: false
show:
in_header: false
- entity: sensor.eppma_2_energy_per_hour
name: 2. energy hrly (utility meter max/hr)
type: area
curve: stepline
yaxis_id: first
stroke_width: 1
extend_to: false
show:
in_header: true
float_precision: 2
opacity: 0.4
- entity: sensor.eppma_3_energy_per_hour_max
name: 3. Max energy/hr
curve: stepline
type: line
yaxis_id: first
stroke_width: 1
extend_to: false
show:
in_header: false
float_precision: 2
- entity: >-
sensor.eppma_4_energy_hourly_value_adjusted_for_reductions_during_non_peak_hours
name: 4. adjusted for non peak hours
type: area
curve: stepline
yaxis_id: first
stroke_width: 3
extend_to: false
show:
in_brush: true
in_header: true
float_precision: 2
opacity: 0.6
- entity: sensor.eppma_5_energy_adjusted_daily_max_peak_value
name: 5. daily max (running)
curve: stepline
yaxis_id: first
stroke_width: 3
extend_to: end
show:
in_brush: true
float_precision: 2
- entity: sensor.eppma_6_energy_peak_per_day
name: 6. peak/d
curve: stepline
yaxis_id: first
stroke_width: 6
extend_to: end
show:
in_brush: true
in_header: false
float_precision: 2
- entity: sensor.eppma_7_peaks_per_month_average_per_day
name: 7. peaks/mth ave this mth
curve: stepline
yaxis_id: first
stroke_width: 8
stroke_dash: 10
extend_to: end
show:
in_brush: false
in_header: true
float_precision: 2
- entity: sensor.eppma_8_peaks_per_month_average_last_months_values
name: 8. peaks/mth last mth
curve: stepline
yaxis_id: first
stroke_width: 3
extend_to: false
show:
in_brush: true
in_header: true
float_precision: 2