Customisation

This document describes the Predbat configuration items in Home Assistant that you can modify to customise Predbat to fit your needs.

All of these settings are entities that can be configured directly in Home Assistant (unlike the 'apps.yaml' configuration items that have to be edited with a file editor).

See Displaying output data for information on how to view and edit these entities within Home Assistant.

Saving and restoring Predbat settings

The selector select.predbat_saverestore can be used to save your current Predbat settings to a yaml file (kept in the directory /config/predbat_save/) and to restore the settings from one of these files.

Selecting the selector option save current will cause the settings to be saved to a date/time stamped file. You can rename this file yourself in the Home Assistant filesystem to give it a more human readable name, or delete it if you no longer want to keep it. This is normally best done in an SSH window or via a Samba mount.

Selecting the option restore default will put all your settings back to the Predbat defaults. Before the restore the current Predbat settings will be saved to the file previous.yaml - should you have made a mistake you can restore them quickly again.

Selecting any of the .yaml files you have created will restore your settings from this file.

Predbat mode

The mode that Predbat operates in will change the operation, this can be configured with select.predbat_mode drop down menu as follows:

Monitor
Control SOC Only
Control charge
Control charge & discharge

If the switch.predbat_set_read_only is set to True then this prevents Predbat from making modifications to the inverter settings (regardless of the configuration). Predbat will continue making and updating its prediction plan every 5 minutes, but no inverter changes will be made. This is useful if you want to over-ride what predbat is planning to do (e.g. your own automation), or whilst you are learning how Predbat works prior to turning it on 'in anger'.

NOTE: Changing the Predbat mode or the read only switch will cause Predbat to reset the inverter settings to default, this will disable both charge and discharge, reset charge and discharge rates to full power and reset the reserve to the default setting

Predbat Monitor mode

In Monitor mode Predbat will not control or Plan any charging or discharging, inverter balancing will take place if enabled, and the plan will show just what is expected based on the current inverter configuration alone.

Predbat Control SOC only mode

In Control SOC only mode Predbat will adjust the target charge percentage (SOC target) according to the Best plan, but the charge window will not be modified.

This mode can be useful if you just have one fixed charge slot per day and you only want Predbat to control the percentage the battery is charged based on solar generation and predicted house load.

CAUTION: You must manually set any charging required on the inverter and if the charge window is disabled then no charging will take place.

Predbat Control charge mode

In Control charge mode Predbat will set the charge times and charge percentages according to the Best plan, charging can be enabled and disabled by Predbat. Predbat will set the inverter into Eco mode when required to enable the battery to support house load, but it will not plan any forced discharging of the battery for export purposes.

This mode can be useful if you don't have an export rate, or if you want to preserve the battery for home demand.

Predbat Control charge & discharge mode

In Control charge & discharge mode Predbat will set both charge and discharge times and control charge and discharge percentages.

If you have set the switch.predbat_set_discharge_freeze_only set to True then forced export won't occur but Predbat can force the export of solar power to the grid when desired.

Expert mode

Predbat has a toggle switch called switch.predbat_expert_mode which is set to Off by default for new installs (On by default for upgraded installs). A lot of Predbat's more advanced configuration options will not be available unless expert mode is enabled. It's recommended for new users to start without expert mode and then maybe enable it later once you become more confident with the tool.

By default Predbat controls the inverter and updates the plan every 5 minutes, this can however use a lot of CPU power especially on more complex tariffs like Agile when run on lower power machines such as Raspberry PIs and some thin clients.

You can tweak input_number.predbat_calculate_plan_every (expert mode) to reduce the frequency of replanning while keeping the inverter control in the 5 minute slots. E.g. a value of 10 or 15 minutes should also give good results.

If you have performance problems leave switch.predbat_calculate_second_pass (expert mode) turned Off as it's quite CPU intensive and provides very little improvement for most systems.

You can can enable combine_charge_slots and combine_discharge_slots in order to speed up planning. Note: Combining discharge slots may prevent optimal forced export. Combining charge slots is usually fine for tariffs with longer periods of fixed rates but can limit the planning ability in some cases.

The number of threads you use can change your performance, you can set threads in apps.yaml to 0 to disable threading if you don't have multiple CPUs available or set it to 'auto' (the default) to use one thread per CPU. Its recommended you don't set this to an odd number of threads.

Battery loss options

input_number.predbat_battery_loss is an assumed percentage figure for energy lost when charging the battery, the default 0.05 is 5%.

input_number.predbat_battery_loss_discharge is an assumed percentage figure for energy lost whilst discharging the battery, the default 0.05 is 5%.

input_number.predbat_inverter_loss is an assumed percentage figure for energy lost during the conversion within the inverter from DC to AC or AC to DC, the default is 0% for legacy reasons but please adjust.

switch.predbat_inverter_hybrid Set to True if you have a hybrid inverter so no inverter losses will be applied for DC charging from Solar generation. Set to False if you have an AC coupled battery and inverter losses will be applied when charging from solar. NB: This switch only applies when Predbat is modelling solar charging. All grid charging (regardless of inverter type) has to undergo an AC to DC conversion and so the inverter_loss % will be included in Predbat's model when charging from the grid.

input_number.predbat_metric_battery_cycle (expert mode) This sets a 'virtual cost' in pence per kWh on using your battery for charging and discharging. Higher numbers will reduce battery cycles at the expense of using higher energy costs.
In theory if you think your battery will last say 6000 complete cycles and cost you £4000 and is 9.5kWh then each full charge and discharge cycle is 19kWh and so the cost per cycle is £4000 / 19 / 6000 = 3.5p.

Taking the 3.5p example, Predbat will apply a "virtual cost" of 3.5p to every kWh of charge and of discharge of the battery. This cost will be included in Predbat's cost optimisation plan when it decides whether to charge, discharge the battery or let the house run on grid import.
NB: For clarity and to re-emphasise, the "virtual cost" will be applied to BOTH the cost calculation for charging AND for discharging the battery.

If you configure this number higher then more expensive plans will be selected which avoids charging and discharging your battery as much. The default is 0.5p (meaning an extra virtual cost of 1p per kWh to charge and discharge) but can be set to 0 if you want to turn this feature off. Note that the cycle cost will not be included in the cost predictions that Predbat produces such as the Predbat HTML plan or Apex charts, its just a cost taken into account by Predbat at the planning stage when the plan is calculated.
NB: Setting this to a non-zero value will increase your daily cost, but will reduce your home battery usage.

Figures of around 0p-2p are recommended, the default is 0.5p per kWh.

input_number.predbat_metric_battery_value_scaling (expert mode) A percentage value that can be used to scale the value of the energy in the battery at the end of the plan. The battery value is accounted for in the optimisations at the lowest future import rate including charging and inverter losses. A value of 1.0 means no change to this, while lower than 1.0 means to value future battery levels less, greater than 1.0 will value it more (and hence hold more charge at the end of the plan).

Scaling and weight options

input_number.predbat_battery_rate_max_scaling is a percentage factor to adjust your maximum charge rate from that reported by the inverter. For example a value of 0.95 would be 95% and indicate charging at 5% slower than reported. For GE inverters the charge rate reports the max AC rate and thus needs to be reduced by inverter losses. You can try computing your discharge curve and check recommendations for changing this figure in the logfile.

input_number.predbat_battery_rate_max_scaling_discharge is a percentage factor to adjust your maximum discharge rate from that reported by the inverter. For GE inverters the discharge rate is reported as the max AC rate and thus is fairly accurate. You can try computing your discharge curve and check recommendations for changing this figure in the logfile.

switch.predbat_battery_capacity_nominal - When enabled Predbat uses the reported battery size from the GivTCP 'Battery Nominal Capacity' field rather than from the normal GivTCP reported 'Battery Capacity kWh' size. If your battery size is reported wrongly maybe try turning this on and see if it helps.

input_number.predbat_load_scaling is a percentage Scaling factor applied to historical load, increase this if you want to be more pessimistic on future consumption. Use 1.0 to use exactly previous load data. A value of 1.1 for example would add 10% to historical load.

input_number.predbat_load_scaling10 is a percentage Scaling factor applied to historical load only for the PV10% scenario (this is in addition to load_scaling above). This can be used to make the PV10% scenario take into account extra load usage and hence be more pessimistic while leaving the central scenario unchanged. The default is 1.1 meaning an extra 10% load is added. This will only have an impact if the PV 10% weighting is non-zero.

input_number.predbat_load_scaling_saving is a percentage Scaling factor applied to historical load only during Octopus Saving sessions. This can be used to model your household cutting down on energy use inside a saving session (e.g. turning off a heat pump, deferring cooking until after the session, etc).

input_number.predbat_pv_scaling is a percentage scaling factor applied to PV data, decrease this if you want to be more pessimistic on PV production vs Solcast.
Use 1.0 to use exactly use the Solcast forecast generation data. A value of 0.9 for example would remove 10% from the Solcast generation forecast.

input_number.predbat_pv_metric10_weight is the percentage weighting given to the Solcast 10% PV scenario in calculating solar generation. Use 0.0 to disable using the PV 10% in Predbat's forecast of solar generation. A value of 0.1 assumes that 1 in every 10 times we will get the Solcast 10% scenario, and 9 in every 10 times we will get the 'median' Solcast forecast.
Predbat estimates solar generation for each half hour slot to be a pv_metric10_weight weighting of the Solcast 10% PV forecast to the Solcast Median forecast.
A value of 0.15 is recommended.

Historical load data

The historical load data is taken from the load sensor as configured in apps.yaml and the days are selected using days_previous and weighted using days_previous_weight in the same configuration file

switch.predbat_load_filter_modal (expert mode) when enabled will automatically discard the lowest daily consumption day from the list of days to use (provided you have more than 1 day selected in days_previous). This can be used to ignore a single low usage day in your average calculation. By default is feature is enabled but can be disabled only in expert mode.

Car Charging

Car charging hold options

Car charging hold is a feature where you try to filter out previous car charging from your historical data so that future predictions are more accurate.

When switch.predbat_car_charging_hold is enabled when for loads of above the power threshold input_number.predbat_car_charging_threshold are assumed to be car charging and input_number.predbat_car_charging_rate will be subtracted from the historical load data.

For more accurate results can you use an incrementing energy sensor set with car_charging_energy in the apps.yaml configuration file. In this case when switch.predbat_car_charging_hold is enabled historical data will be subtracted from the load data instead of using the fixed threshold method.

input_number.predbat_car_charging_energy_scale Is used to scale the car_charging_energy sensor, the default units are kWh so if you had a sensor in watts you might use 0.001 instead.

input_number.predbat_car_charging_rate - Set to the car's charging rate in kW per hour (normally 7.5 for 7.5kWh), but will be pulled automatically from Octopus Energy integration if enabled for Octopus Intelligent.

input_number.predbat_car_charging_loss gives the amount of energy lost when charging the car (load in the home vs energy added to the battery). A good setting is 0.08 which is 8%.

Car charging plan options

Car charging planning - is only used if Intelligent Octopus isn't enabled and car_charging_planned is connected correctly.

This feature allows Predbat to create a plan for when you car will charge, but you will have to create an automation to trigger your car to charge using binary_sensor.predbat_car_charging_slot if you want it to match the plan.

select.predbat_car_charging_plan_time - When using Predbat-led planning set this to the time you want the car to be charged by
switch.predbat_car_charging_plan_smart - When enabled (True) allows Predbat to allocate car charging slots to the cheapest times, when disabled (False) all low rate slots will be used in time order.
input_number.predbat_car_charging_plan_max_price - When non-zero sets a maximum price per kWh to pay when charging your car, when disabled (0) all slots will be considered.

switch.predbat_octopus_intelligent_charging when true enables the Intelligent Octopus charging feature which will make Predbat create a car charging plan which is taken from the Intelligent Octopus plan you must have set the octopus_intelligent_slot sensor in apps.yaml to enable this feature.

If Octopus Intelligent Charging is enabled the switch switch.predbat_octopus_intelligent_ignore_unplugged (expert mode) can be used to prevent Predbat from assuming the car will be charging when the car is unplugged. This will only work correctly if car_charging_planned is set correctly in apps.yaml to detect your car being plugged in.

Control how your battery behaves during car charging:

switch.predbat_car_charging_from_battery - When True the car can drain the home battery, Predbat will manage the correct level of battery accordingly. When False home battery discharge will be prevented when your car charges, all load from the car and home will be from the grid. This is achieved by setting the discharge rate to 0 during car charging and to the maximum otherwise, hence if you turn this switch Off you won't be able to change your discharge rate outside Predbat. The home battery can still charge from the grid/solar in either case. Only use this if Predbat knows your car charging plan, e.g. you are using Intelligent Octopus or you use the car slots in Predbat to control your car charging.

If your car does not have an SOC sensor and you are not using Octopus Intelligent you can set switch.predbat_car_charging_manual_soc to have Predbat create input_number.predbat_car_charging_manual_soc_kwh which will hold the cars current state of charge (soc) in kWh. You will need to manually set this to the cars current charge level before charging, Predbat will increment it during charging sessions but will not reset it automatically.

Calculation options

See the Predbat mode setting as above for basic calculation options

input_number.predbat_forecast_plan_hours is the minimum length of the Predbat charge plan, and is the number of hours after the first charge slot to include in the plan. The default of 24 hours is the recommended value (to match energy rate cycles). Note that the actual length of the Predbat plan will vary depending upon when the first charge slot is.

switch.predbat_calculate_regions (expert mode) When True the a second pass of the initial thresholds is calculated in 4 hour regions before forming the detailed plan. Is True by default but can be turned off in expert mode.

switch.predbat_calculate_discharge_oncharge (expert mode) When True calculated discharge slots will disable or move charge slots, allowing them to intermix. When False discharge slots will never be placed into charge slots.

switch.predbat_set_discharge_during_charge - If turned off disables inverter discharge during charge slots, useful for multi-inverter setups to avoid cross charging when batteries are out of balance.

switch.predbat_calculate_tweak_plan (expert mode) When True causes Predbat to perform a second pass optimisation across the next 8 charge and discharge windows in time order.

This can help to slightly improve the plan for tariffs like Agile but can make it worse in some fixed rate tariffs which you want to discharge late.

switch.predbat_calculate_second_pass (expert mode) When True causes Predbat to perform a second pass optimisation across all the charge and discharge windows in time order.

NOTE: This feature is quite slow and so may need a higher performance machine.

This can help to slightly improve the plan for tariffs like Agile but can make it worse in some fixed rate tariffs which you want to discharge late.

Battery margins and metrics options

input_number.predbat_best_soc_keep is the minimum battery level in kWh that Predbat will to try to keep the battery above for the Predbat plan. This is a soft constraint only that's used for longer term planning and is ignored for the forthcoming first 4 hours of the plan. As this is not used for short-term planning it's possible for your SoC to drop below this - use input_number.predbat_best_soc_min if you need a hard SoC constraint that will always be maintained. It's usually good to have best_soc_keep set to a value above 0 to allow some margin in case you use more energy than planned between charge slots.

input_number.predbat_best_soc_min (expert mode) sets the minimum charge level (in kWh) for charging during each slot and the minimum discharge level also (set to 0 if you want to skip some slots). If you set this to a non-zero value you will need to use the low rate threshold to control which slots you charge from or you may charge all the time.

input_number.predbat_best_soc_max (expert mode) sets the maximum charge level (in kWh) for charging during each slot. A value of 0 disables this feature.

input_number.combine_rate_threshold (expert mode) sets a threshold (in pence) to combine charge or discharge slots together into a single larger average rate slot. The default is 0p which disables this feature and all rate changes result in a new slot.

switch.predbat_combine_charge_slots Controls if charge slots of > 30 minutes can be combined. When disabled they will be split up, increasing run times but potentially more accurate for planning. Turn this off if you want to enable ad-hoc import during long periods of higher rates but you wouldn't charge normally in that period (e.g. pre-charge at day rate before a saving session). The default is disabled (False)

switch.predbat_combine_discharge_slots (expert mode) Controls if discharge slots of > 30 minute can be combined. When disabled they will be split up, increasing run times but potentially more accurate for planning. The default is disabled (False)

input_number.predbat_metric_min_improvement (expert mode) sets the minimum cost improvement in pence that it's worth lowering the battery SOC % for. The default value is 0 which means this feature is disabled and the battery will be charged less if it's cost neutral. If you use input_number.predbat_pv_metric10_weight then you probably don't need to enable this as the 10% forecast does the same thing better Do not use if you have multiple charge windows in a given period as it won't lead to good results (e.g. Agile) You could even go to something like -0.1 to say you would charge less even if it cost up to 0.1p more (best used with metric10).

input_number.predbat_metric_min_improvement_discharge (expert mode) Sets the minimum pence cost improvement it's worth doing a forced discharge (and export) for. A value of 0.1 is the default which prevents any marginal discharges. If you increase this value (e.g. you only want to discharge/forced export if definitely very profitable), then discharges will become less common and shorter. The value is in pence per 30 minutes of export time.

input_number.predbat_rate_low_threshold (expert mode) When set to 0 (the default) Predbat will automatically look at the future import rates in the plan and determine the import rate threshold below which a slot will be considered to be a potential charging slot.
If rate_low_threshold is set to a non zero value this will set the threshold below future average import rates as the minimum to consider for a charge window, e.g. setting to 0.8 = 80% of average rate.
If you set this too low you might not get enough charge slots. If it's too high you might get too many in the 24-hour period which makes optimisation harder.

input_number.predbat_rate_high_threshold (expert mode) When set to 0 (the default) Predbat will automatically look at the future export rates in the plan and determine the threshold above which a slot can be considered a potential exporting slot.
If rate_high_threshold is set to a non zero value this will set the threshold above future average export rates as the minimum export rate to consider exporting for, e.g. setting to 1.2 = 20% above average rate.
If you set this too high you might not get any export slots. If it's too low you might get too many in the 24-hour period.

input_number.predbat_metric_future_rate_offset_import (expert mode) Sets an offset to apply to future import energy rates that are not yet published, best used for variable rate tariffs such as Agile import where the rates are not published until 4pm. If you set this to a positive value then Predbat will assume unpublished import rates are higher by the given amount.

Setting this to 1 to 1.5p for example results in Predbat being a little more aggressive in the charging calculation for today - Predbat will charge the battery to a higher percentage than it would otherwise as it expects a cost benefit of using today's lower rates. NB: this can lead to higher costs and to some export if solar generation is better than forecast.

input_number.predbat_metric_future_rate_offset_export (expert mode) Sets an offset to apply to future export energy rates that are not yet published, best used for variable rate tariffs such as Agile export where the rates are not published until 4pm. If you set this to a negative value then Predbat will assume unpublished export rates are lower by the given amount.

switch.predbat_calculate_inday_adjustment (expert mode) Enabled by default with damping of 0.95. When enabled will calculate the difference between today's actual load and today's predicated load and adjust the rest of the days usage prediction accordingly. A scale factor can be set with input_number.predbat_metric_inday_adjust_damping (expert mode) to either scale up or down the impact of the in-day adjustment (lower numbers scale down its impact). The in-day adjustment factor can be seen in predbat.load_inday_adjustment and charted with the In Day Adjustment chart (template can be found in the charts template in Github).

input_number.predbat_carbon_metric (carbon enable) When Carbon footprint tracking is enabled (switch.predbat_carbon_enable) you can specify a cost per Kg of CO2 used to weight the selection of plans. Values of around 10-200 will give varying outcomes to trade off cost vs carbon footprint of your system.

Note: Carbon footprint tracking can only be enabled if apps.yaml is configured to point to the correct CO2 cost sensor

Inverter control options

switch.predbat_set_status_notify Enables mobile notification about changes to the Predbat state (e.g. Charge, Discharge etc). On by default.

switch.predbat_set_inverter_notify Enables mobile notification about all changes to inverter registers (e.g. setting window, turning discharge on/off). Off by default.

switch.predbat_set_charge_low_power Enables low power charging mode where the max charge rate will be automatically determined by Predbat to be the lowest possible rate to meet the charge target. This is only really effective for charge windows >30 minutes. If this setting is turned on, its strongly recommended that you create a battery_power_charge_curve in apps.yaml as otherwise the low power charge may not reach the charge target in time. This setting is off by default.

The YouTube video low power charging and charging curve explains how the low power charging works and shows how Predbat automatically creates it.

switch.predbat_set_reserve_enable (expert_mode) When enabled the reserve setting is used to hold the battery charge level once it has been reached or to protect against discharging beyond the set limit. Enabled by default.

switch.predbat_set_charge_freeze (expert mode) When enabled will allow Predbat to hold the current battery level while drawing from the grid/solar as an alternative to charging. Enabled by default.

switch.predbat_set_discharge_freeze_only (expert mode) When enabled forced discharge is prevented, but discharge freeze can be used (if enabled) to export excess solar rather than charging the battery. This is useful with tariffs that pay you for solar exports but don't allow forced export (brown energy).

If you have switch.predbat_inverter_hybrid set to False then if switch.predbat_inverter_soc_reset (expert mode) is set to True then the target SOC % will be reset to 100% outside a charge window. This may be required for AIO inverter to ensure it charges from solar. The default for this switch is True but it can be disabled in expert mode if need be.

input_number.predbat_set_reserve_min Defines the reserve percentage to reset the reserve to when not in use, a value of 4 is the minimum and recommended to make use of the full battery.
If you want to pre-prepare the battery to retain extra charge in the event of a high likelihood of a grid power outage such as storms predicted, you can increase set_reserve_min to 100%, and then change it back afterwards.
(Obviously this is only any use if your inverter is wired to act as an Emergency Power Supply or whole-home backup 'island mode' on the GivEnergy AIO).

switch.predbat_inverter_soc_reset (expert mode) When enabled the target SOC for the inverter(s) will be reset to 100% when a charge slot is not active, this can be used to workaround some firmware issues where the SOC target is used for solar charging as well as grid charging. When disabled the SOC % will not be changed after a charge slot. This is disabled by default.

Balance Inverters

When you have two or more inverters it's possible they get out of sync so they are at different charge levels or they start to cross-charge (one discharges into another). When enabled, balance inverters tries to recover this situation by disabling either charging or discharging from one of the batteries until they re-align.

The apps.yaml contains a setting balance_inverters_seconds which defines how often to run the balancing, 30 seconds is recommended if your machine is fast enough, but the default is 60 seconds.

Enable the switch.predbat_balance_inverters_enable switch in Home Assistant to enable this feature.

switch.predbat_balance_inverters_charge - Is used to toggle on/off balancing while the batteries are charging
switch.predbat_balance_inverters_discharge - Is used to toggle on/off balancing while the batteries are discharging
switch.predbat_balance_inverters_crosscharge - Is used to toggle on/off balancing when the batteries are cross charging
input_number.predbat_balance_inverters_threshold_charge - Sets the minimum percentage divergence of SoC during charge before balancing, default is 1%
input_number.predbat_balance_inverters_threshold_discharge - Sets the minimum percentage divergence of SoC during discharge before balancing, default is 1%

Cloud coverage and load variance

Predbat tries to model passing clouds by modulating the PV forecast data on a 5 minute interval up and down while retaining the same predicted total. The amount of modulation depends on the difference between the PV50% (default) and PV10% scenario produced by Solcast.

You can disable this feature (expert mode only) using switch.predbat_metric_cloud_enable

Predbat tries to model changes in your household load by modulating the historical data on a 5 minute interval up and down while retaining the same predicted total. The amount of modulation depends on the standard deviation of your load predictions over the coming period (currently 4 hours).

You can disable this feature (expert mode only) using switch.metric_load_divergence_enable

iBoost model options

Predbat has an 'iBoost model' that can be used to model using excess solar energy to heat hot water (or similar) instead of it being exported to the grid.

This model can be used to control any solar diverter device, for example an iBoost (e.g. using a Fingerbot or similar device to physically press the 'boost' button on the iBoost), a MyEnergy Eddi (using the MyEnergy integration), or it can be used with a high power smart switch to turn on the hot water cylinder immersion heater when there is excess solar.

So although Predbat refers to controlling an iBoost, you are not limited to just an iBoost device when using this model within Predbat.

To turn the model on, switch.predbat_iboost_enable needs to be enabled.

The predicted output from the iBoost solar diverter model is returned in predbat.iboost_best and is populated in the 'iBoost' column of the Predbat plan.

When you turn on predbat_iBoost_enable the following additional Home Assistant entities are created by Predbat:

switch.predbat_iboost_solar When enabled assumes the solar diverter will use solar power to boost the hot water heating.
input_number.predbat_iboost_min_soc sets the minimum home battery SoC percentage that must be in the battery before the solar diverter is turned on. The default is 0 meaning hot water heating can occur regardless of what SoC level the battery is at.
switch.predbat_iboost_gas When enabled will control the solar diverter to only operate when electric rates are lower than gas rates. This is useful if you have the choice to heat your hot water by immersion heater or by gas boiler. Note: Gas rates have to be configured in apps.yaml using metric_octopus_gas.
input_number.predbat_iboost_gas_scale Sets the scaling of the gas rates (set in apps.yaml)used before comparing with electric rates, to account for gas boiler losses and efficiency. It should be set to the reciprocal of the boiler efficiency, i.e. for an 80% efficient gas boiler, set to 1.25.
switch.predbat_iboost_charging If set to on, the solar diverter will only operate when the battery is charging (can be combined with iboost_gas or not).
input_number.predbat_iboost_max_energy Sets the maximum energy in kWh that the solar diverter can consume during a day before turning off - default 3kWh.
input_number.predbat_iboost_max_power Sets the maximum power in watts that the solar diverter will consume - default 2400.
input_number.predbat_iboost_min_power Sets the minimum power in watts that the solar diverter will consume - default 500.

You will see input_number.predbat_iboost_today entity which tracks the estimated kWh consumed by the solar diverter during the day, and resets at 11:30pm every night.

The binary_sensor.predbat_iboost_active entity will be enabled when the solar diverter should be active, can be used for automations to trigger the immersion heater boost.

Example template automation for controlling the solar diverter:

alias: Solar Diverter
description: "Start/stop solar diverter based on Predbat determined slots"
trigger:
  - platform: state
    entity_id:
     - binary_sensor.predbat_iboost_active
action:
  - choose:
      - conditions:
          - condition: state
            entity_id: binary_sensor.predbat_iboost_active
            state: "True"
        sequence:
          <commands to turn on your solar diverter>
      - conditions:
          - condition: state
            entity_id: binary_sensor.predbat_iboost_active
            state: "False"
        sequence:
          <commands to turn off your solar diverter>
mode: single

If you have an incrementing sensor that tracks the solar diverter energy usage then to make your predictions more accurate you should set the iboost_energy_today sensor in apps.yaml to point to it, and optionally set iboost_energy_scaling if the sensor isn't in kWh (e.g. set to 0.001 if the sensor is in Watts). The sensor should be an incrementing sensor which can reset at midnight or not.

Holiday mode

When you go away you are likely to use less electricity and so the previous load data will be quite pessimistic.

Using the Home Assistant entity input_number.predbat_holiday_days_left you can set the number of full days that you will be away for (including today). The number will count down by 1 day at midnight until it gets back to zero. Whilst holiday days left is non-zero, Predbat's 'holiday mode' is active.

When Predbat's 'holiday mode' is active the historical load data will be taken from yesterday's data (1 day ago) rather than from the days_previous setting in apps.yaml. This means Predbat will adjust more quickly to the new usage pattern.

If you have been away for a longer period of time (more than your normal days_previous setting) then obviously it's going to take longer for the historical data to catch up, you could then enable holiday mode for another 7 days after your return.

In summary:

For short holidays set holiday_days_left to the number of full days you are away, including today but excluding the return day
For longer holidays set holiday_days_left to the number of days you are away plus another 7 days until the data catches back up

Manual control

In some cases you may want to override Predbat's planned behaviour and make a decision yourself. One way to achieve this is to put Predbat into read-only mode using switch.predbat_set_read_only. When going to read only mode the inverter will be put back to the default settings and you should then control it yourself using GivTCP or the App appropriate to your inverter.

A better alternative in some cases is to tell Predbat what you want it to do using the manual force features:

You can force the battery to be charged within a 30 minute slot by using the select.predbat_manual_charge selector. Pick the 30 minute slot you wish to charge in, and Predbat will change the plan to charge in the selected slot. You can select multiple slots by using the drop down menu more than once. When Predbat updates the plan you will see the slots picked to be charging slots in the current value of this selector, and annotated in the Predbat HTML plan with an upside down 'F' symbol.

You can cancel a force slot by selecting the slot time again (it will be shown in square brackets to indicate its already selected).

The select.predbat_manual_discharge selector can be used to manually force a discharge within a 30 minute slot in the same way as the manual force charge feature. The force discharge takes priority over force charging.

The select.predbat_manual_idle selector is used to force Predbat to idle mode during a 30 minute slot, this implies no forced grid charging or discharging of the battery. House load will be supplied from solar, or from the battery if there is insufficient solar, or grid import if there is insufficient battery charge. This is described as 'ECO' Mode for GivEnergy inverters but other inverters use different terminology.

The select.predbat_manual_freeze_charge selector is used to force Predbat to freeze charge during a 30 minute slot, this implies the battery will not discharge and hold at the current level. The grid maybe used if solar is not enough to cover the load.

The select.predbat_manual_freeze_discharge selector is used to force Predbat to freeze discharge during a 30 minute slot, this implies the battery will not charge but will still discharge for the house load. Any solar will be exported to the grid.

When you use the manual override features you can only select times in the next 18 hours, the overrides will be removed once their time slot expires (they do not repeat).

_NOTE: once you select a time slot from any of the select.predbat_manual_ selectors the selected time slot is immediately marked on the drop-down and you can then make another change. Predbat still has to update the plan which it will be doing so in the background, and this can take a few minutes to run (depending on the speed and power of the PC you are running Home Assistant on) so don't be surprised why the Predbat plan doesn't change immediately - remember you can see the date/time the plan was last updated on the first row of the plan.

CAUTION: If you leave Predbat turned off for a long period of time then the override timeslots could end up repeating when you restart

Debug

switch.predbat_debug_enable when on prints lots of debug, leave off by default

switch.predbat_plan_debug (expert mode) when enabled adds some extra debug to the Predbat HTML plan - see Predbat Plan debug mode for more details.