@@ -27,6 +27,109 @@ To set up the integration, you need a username and password with access to the [
2727
2828{% include integrations/config_flow.md %}
2929
30+ ## Examples
31+
32+ ### Statistics graph
33+
34+ An example of a Statistics graph that shows hourly production per module for the past 7 days.
35+
36+ ![ Screenshot hourly production Statistics graph] ( /images/integrations/solaredge_modules/hourly_production.png )
37+
38+ ``` yaml
39+ chart_type : line
40+ period : hour
41+ type : statistics-graph
42+ entities :
43+ - solaredge_modules:1234567_100000001
44+ - solaredge_modules:1234567_100000002
45+ - solaredge_modules:1234567_100000003
46+ - solaredge_modules:1234567_100000004
47+ stat_types :
48+ - change
49+ days_to_show : 7
50+ title : Hourly production per module on east side
51+ ` ` `
52+
53+ Another example of a Statistics graph that shows daily production per module for the past 7 days.
54+ It's easier here to identify any problematic modules.
55+
56+ 
57+
58+ ` ` ` yaml
59+ chart_type : line
60+ period : day
61+ type : statistics-graph
62+ entities :
63+ - solaredge_modules:1234567_100000001
64+ - solaredge_modules:1234567_100000002
65+ - solaredge_modules:1234567_100000003
66+ - solaredge_modules:1234567_100000004
67+ stat_types :
68+ - change
69+ title : Daily production per module on east side
70+ days_to_show : 30
71+ ` ` `
72+ ### SQL
73+
74+ To identify problematic modules you could setup the [` SQL`](/integrations/sql/) integration with the following:
75+
76+ Name : SolarEdge low production modules (East)
77+
78+ Column : ` problematic_modules`
79+
80+ Query :
81+
82+ ` ` ` sql
83+ SELECT * FROM (
84+ WITH RelevantTimeRange AS (
85+ -- Define the start and end timestamps for "past 7 days, ignoring today"
86+ -- start_ts is inclusive, end_ts is exclusive
87+ SELECT
88+ strftime('%s', date('now', 'localtime', 'start of day', '-7 days')) AS period_start_ts,
89+ strftime('%s', date('now', 'localtime', 'start of day')) AS period_end_ts
90+ ),
91+ ModuleProduction AS (
92+ -- Calculate total production for each module in the relevant time period
93+ SELECT
94+ sm.name,
95+ SUM(s.state) AS total_production
96+ FROM statistics s
97+ JOIN statistics_meta sm ON s.metadata_id = sm.id
98+ CROSS JOIN RelevantTimeRange rt
99+ WHERE
100+ sm.source = 'solaredge_modules'
101+ -- TODO: Adjust this to match your setup
102+ AND sm.name LIKE '% 1.1.%'
103+ AND s.start_ts >= rt.period_start_ts
104+ AND s.start_ts < rt.period_end_ts
105+ AND s.state IS NOT NULL
106+ GROUP BY
107+ sm.name
108+ ),
109+ AverageProduction AS (
110+ -- Calculate the average production across all modules that produced something
111+ SELECT
112+ AVG(mp.total_production) AS average_total_production
113+ FROM ModuleProduction mp
114+ WHERE total_production > 0
115+ )
116+ SELECT
117+ GROUP_CONCAT(
118+ mp.name || ' (' || printf("%.1f", (mp.total_production * 100.0 / ap.average_total_production)) || '%)',
119+ ', '
120+ ) AS problematic_modules
121+ FROM
122+ ModuleProduction mp, AverageProduction ap -- Implicit cross join, AP will have 1 row
123+ WHERE
124+ -- TODO: Adjust the 96% threshold if needed
125+ mp.total_production < (0.96 * ap.average_total_production)
126+ AND ap.average_total_production > 0 -- Avoid division by zero if no production at all
127+ )
128+ ` ` `
129+
130+ This will result to a sensor with state e.g. : ` SolarEdge 1.1.13 (95.7%), SolarEdge 1.1.14 (95.2%)`
131+ You can use this sensor in automations, e.g to notify you if its value changes.
132+
30133# # Removing the integration
31134
32135{% include integrations/remove_device_service.md %}
0 commit comments