Storage, EV & Specialized Systems

Where these systems live

Everything in this chapter is a card. The Project tab (/project) lays its cards out as a responsive masonry — CLIENT, SIMULATION, BATTERY, CONSUMPTION RATES, GRID PRICING, FINANCIAL, CARBON FOOTPRINT, TENDER and Systems — and the specialized systems all live inside that last one. The Systems card holds up to four blocks, each with its own heading: Backup generator (weak grid), Off-grid system, Solar pumping and DC traction grid.

Two more pieces of this chapter sit elsewhere, because they belong to another part of the workflow:

• the electric vehicle is a consumption source, so you add it on the Consumption tab (/consumption) — see §Electric-vehicle co-simulation;

• the grid-availability schedules are the time-varying terms behind two loss-cascade rows, so they live in the Schedules & grid card on the Production tab (/production) — see §Grid-availability schedules. The one exception is the injection cap, a rule of the grid connection that sits on the GRID PRICING card.

The System type gate

Almost every card in this chapter is gated by one control: System type, the first row of the SIMULATION card. It is a dropdown grouped under three headings — Grid-connected, Off-grid and DC grid:

The selector holds exactly these six modes. A project that pins no mode at all opens on Self-consumption + storage.

Which card appears, then, is decided as follows:

• BATTERY — Self-consumption + storage or Off-grid site.

• Backup generator (weak grid) — Self-consumption + storage and the battery Dispatch strategy set to Weak-grid backup. Both conditions, because the engine only dispatches this generator in that one arrangement.

• Off-grid system — follows its own Off-grid (no grid connection) checkbox, which picking Off-grid site arms for you. Unticking the checkbox moves the selector back to Self-consumption + storage, so the mode and the card can never disagree.

• Solar pumping — Off-grid solar pumping only.

• DC traction grid — DC grid (rail / trolleybus) only.

• The Systems card itself disappears when the active mode unlocks none of the four, rather than standing empty.

Battery storage

The BATTERY card has two parts: the Dispatch strategy that decides how the battery is operated, and the bank list below it that decides which catalogue batteries make up the storage.

Dispatch strategy

At the top of the card, Dispatch strategy selects one of four dispatch algorithms. The strategy decides, slot by slot, when the battery charges and when it discharges.

Peak shaving and Time of use both read their power from Peak-shave target (in kW) on the GRID PRICING card (The project's consumption-rate and grid-pricing cards). That row appears whenever the strategy is Peak shaving, Time of use or Weak-grid backup, whatever the card's Granularity, and at TURPE + peak-shaving or Net metering whatever the strategy. Left at 0, both strategies fall back to Self-consumption. Weak-grid backup uses the same figure as its discharge cap during an outage, or the pack's own capacity per hour when it is 0.

Choosing Weak-grid backup is also what reveals the Backup generator (weak grid) block (§Hybrid diesel / genset generator), so the two are configured together.

Building the battery bank

Below the strategy sits the bank list. It always has a fixed Primary line and lets you add more, captioned Bank 1, Bank 2 and so on. Each line is a Battery from the catalogue dropdown and a quantity.

Dropdown entries are prefixed with the battery's kind — [Phys.] for a physical pack, [Virt.] for a virtual bank — followed by the manufacturer and model. None is always the first entry and means “no battery on this line”. The last entry, Add from the catalogue…, is not a battery: it opens the catalogue on its Physical batteries tab, where hearting a battery adds it to these dropdowns. The line keeps the battery it had.

Virtual and physical batteries

pvSpline models two very different kinds of storage, and which one you get is decided by the catalogue entry you pick, not by a switch on the card.

Picking a physical battery carries over its capacity, maximum charge and discharge power, round-trip efficiency, minimum state-of-charge (depth-of-discharge floor), cycle life, calendar life, and — when the catalogue entry has one — its install cost, which becomes the replacement cost the financial model charges at end of life.

Picking a virtual battery carries over its credited capacity (a catalogue capacity of zero means “no cap”), the withdrawal fee per kWh, the flat monthly subscription, the per-kWp monthly subscription, the deposit conversion ratio, the billing period (monthly or annual) and whether unused credits expire at the end of each period.

Aging, warranty and replacement

Physical batteries lose capacity over time, and pvSpline models this so the financial table reflects a shrinking usable pack. The cycle and calendar endpoints come from the catalogue entry; the engine takes the worse of the two, applies a warranty floor, and — when a replacement cost is known — triggers an automatic replacement in the first year the modelled state of health drops below the end-of-life floor, charging the cost as a one-time outflow and restarting the aging clock. Without a replacement cost the pack simply stays clamped at end-of-life for the rest of the horizon.

The underlying models (a conservative linear worse-of-calendar-or-cycle fade, and a BLAST-Lite physical model using square-root-of-time calendar fade and square-root-of-throughput cycle fade with depth-of-discharge scaling and an Arrhenius temperature term) are described in Simulation Physics & Data Reference. A project that already carries those settings keeps them exactly as stored.

When the Primary line is on None and there is no extra bank, there is no pack to age: the run replaces nothing and charges no replacement cost.

Sizing the battery

Three tools help you pick a battery instead of guessing.

Catalogue comparison. On the Batch tab (/batch), the Parts page with the Batteries category runs one full simulation per battery you tick and ranks them side by side on NPV, LCOE and the rest. Add a scenario that keeps the project as it is, and the comparison also tells you whether a battery pays at all. See Batch Simulation & Scenario Sweeps.

Joint sizing. The same tab's Auto-size page can pick the battery while it sizes the array: tick Sweep batteries and every power in the candidate grid is tried with the project's own battery and with each catalogue battery you favourited, each candidate costed on its real battery spend. It needs an Install rate — without one the page warns that the part axes will be skipped — and the checkbox is greyed out while the project simulates no storage. Applying the winner — with Apply to project, or in one click with Optimize now — makes its battery the project's battery. See Batch Simulation & Scenario Sweeps.

Off-grid pre-sizing. For stand-alone sites, the pre-sizing page (§Pre-sizing aids) suggests a PV power and a battery capacity from your autonomy and reliability targets.

Electric-vehicle co-simulation

An electric vehicle is added on the Consumption tab (/consumption): click + on the source tab strip to add a source, then set its Type to Electric vehicle. Unlike a fixed load, the EV is co-simulated — the engine decides when to charge it (and, optionally, when to draw energy back from it), so it interacts with the PV surplus and with the stationary battery.

The vehicle, stated as a sentence

The card (EV charging parameters) reads as plain sentences, and every highlighted word in them is the actual control for the value it names.

“My Tesla Model 3 (LR) gets home at 18:00 and leaves again at 07:00.”

The vehicle token is a combobox. Pick a model from its arrow to pre-fill the pack capacity, the onboard AC charger power and the WLTP consumption — ten common models are listed (Tesla Model 3 LR and Model Y LR, Renault Zoe R135 and Megane E-Tech, Peugeot e-208 and e-308, Volkswagen ID.3 Pro and ID.4, Hyundai Kona Electric 64, BMW i4 eDrive40). Picking a model also reseeds the arriving charge to half the new pack, because “25 kWh” means something different on a 50 kWh car and an 80 kWh one. Choose Custom instead and the token turns into a text box where you name the car yourself; that name is the source's label, so it also titles the source's tab.

The two times are the plug-in and unplug times, in 15-minute steps. Only while plugged in can the car charge (or, with lend-back, discharge). The unplug time may cross midnight — the default 18:00 → 07:00 is an overnight window.

“By then it needs 10 kWh this session, charged on sunshine only.”

The number is the charge need; the words after it are the dropdown that says what the number means:

Switching between a kWh reading and the percentage reading converts the value against the pack capacity, so “30 kWh by departure” becomes 60 % on a 50 kWh pack rather than a bare 30 %.

The last token is the charging strategy, which folds the charging policy and the lend-back decision into one choice:

Weekends. Tick Weekends are different to add a second sentence with a separate Saturday/Sunday plug-in window (it arms at 09:00 → 18:00). The two weekend times are a pair: arming or clearing the checkbox sets or clears both, because a lone half is ignored. In the km of driving mode a separate It also drives a different distance at weekends checkbox arms a distinct weekend distance (60 km when first armed); leave it off and the weekday distance is driven every day.

The picture and the verdict

Under the sentences, a diagram shows the house and its roof, the grid, the wallbox and the car, with the pack drawn as one bar: the charge it arrives with, the slice this session adds, and the slice it cannot reach. The arrows show where the kWh come from — the roof for solar-only charging, the grid for the cheap-tariff strategy — and an arrow back to the house when lend-back is on. It is also an editor:

• drag the round knob to set the charge the car arrives with (clamped to the engine's 20 %-of-pack floor);

• drag the flag to set the charge wanted at departure — this also switches the need mode to kWh in the pack by departure, because that is exactly what the flag states;

• press the wallbox to step the charger through the common AC sizes 3.7, 7.4, 11 and 22 kW.

Below the picture, a verdict box answers “will this work?” from the session alone — no simulation needed, because the window's length against what the charger can move through it is pure arithmetic:

The numbers behind it

A The numbers behind it drawer holds the typed surface for the same values: EV battery (pack capacity), Charger (wallbox power), Initial charge (the charge when the car first gets home — later arrivals follow from the charging, any lending to the house and the trips) and, outside the km/day mode, Consumption in Wh/km. Choosing Custom on the model combobox opens the drawer, since those three numbers are what “custom” means. Editing any of them snaps the model back to Custom, so the label never lies about the specs.

Several EV sources can coexist — add one per car and each gets its own tab in the source strip.

Off-grid (stand-alone) site

Choosing Off-grid site reveals the Off-grid system block on the Systems card. Here there is no grid to fall back on: whatever PV, battery and generator cannot cover becomes missing energy, quantified as a loss-of-load probability (LOL).

Reliability parameters

Off-grid (no grid connection) is the block's master checkbox: it enables off-grid mode, with no grid injection or withdrawal. It is armed for you when you pick the mode, and unticking it returns the project to Self-consumption + storage.

Deep-discharge floor is the protected state of charge — the battery never drops below this percentage of capacity, preserving its life on a site with no grid safety net. Autonomy target (pre-sizing) and Target loss of load (pre-sizing) record your sizing targets with the project: how many sunless days the battery should cover from full charge, and the acceptable fraction of annual demand left unserved. The simulation does not use them, and the pre-sizing page (§Pre-sizing aids) has its own Battery autonomy and Target loss-of-load (0–1) fields rather than reading these. Value of lost load monetises any shortfall in the financial balance; leave it at 0 to report reliability without pricing it.

Backup generator (off-grid)

Tick Backup genset to add a diesel generator that takes over when PV + battery run short. Its parameters — Genset rated power, Genset fuel (litres per kWh produced), Genset fuel price — feed the financial OPEX. If Genset recharges the battery is on, the genset runs in set-point mode: it starts when the state of charge drops below Genset start SOC and runs until the battery is recharged to Genset stop SOC (fewer starts, more fuel). Otherwise it merely follows the load.

Hybrid diesel / genset generator

The Backup generator (weak grid) block adds a diesel generator as a third backup source, in the priority order PV → battery → generator → uncovered load. It is the detailed counterpart to the off-grid in-block genset, and it appears only on a grid-tied Self-consumption + storage project whose battery Dispatch strategy is Weak-grid backup — the note at the foot of the block says so: “the generator runs only when the dispatch mode is set to weak-grid backup”.

Enabled is the master checkbox. Left off, there is no generator and behaviour is unchanged.

Choosing a model

Model / template pre-fills the rating and fuel curve. It offers Custom (keep whatever is typed), four generic diesel presets — Generic diesel 8 kVA (6.4 kW), Generic diesel 30 kVA (24 kW), Generic diesel 100 kVA (80 kW) and Generic diesel 250 kVA (200 kW) — and every generator in the catalogue, then, last, Add from the catalogue…, which opens the catalogue's Generators tab instead of filling anything. A pick fills only the datasheet fields (rating, fuel coefficients, minimum load, emission factor); the economics and the SOC thresholds stay yours to set.

Sizing & control

Rated power is the generator's electrical rating. Control strategy chooses how it runs:

Minimum load enforces a floor — a wet-stacking guard; surplus above the served load recharges the battery. Start SOC threshold and Stop SOC threshold bound the cycle-charging band, and Minimum run time keeps the genset running for a minimum period after each start, so start counts stay realistic.

Fuel model & emissions

The fuel curve is the two-coefficient HOMER form Fuel (L/h) = F₀ · P_rated + F₁ · P_out. No-load fuel F0 is the idling term (~0.08 L/h/kW typical) and Marginal fuel F1 the per-kWh term, so the specific consumption at full load is roughly F₀ + F₁ ≈ 0.30 L/kWh. CO₂ sets the combustion emission factor (diesel ~2.68 kg/L; adjust for gas or biofuel), feeding the carbon balance.

Economics

CAPEX folds into the installation cost (depreciated, financed, included in LCOE). Fuel price times the litres burned gives the operating cost, escalated annually by Fuel escalation independently of general inflation. Maintenance is charged per running hour, and Overhaul cost at Overhaul year models a one-time mid-life major service; either set to 0 disables it.

Solar pumping

Choosing Off-grid solar pumping reveals the Solar pumping block. There is no electrical load and no tariff: the “load” is a daily water volume, and the objective is to meet it from the sun, buffering water in a tank. The consumption-rate, grid-pricing and financial cards disappear with the mode, because grid economics do not apply.

Hydraulic circuit

Hydraulic circuit sets the site type, which decides which head parameters matter:

Static head is the resting water level or geometric lift; Friction k_f is the quadratic head-loss coefficient (H_f = k_f · Q²). All the head rows stay on screen whichever circuit you choose, so a stored value is never hidden from you — only some of them are read by the physics.

Pump and power conditioning

Pump technology is either Centrifugal (low head, high flow) or Positive displacement (high head, flow proportional to speed). The pump is described by Pump nominal power, Max flow (flow at zero head, the pump-curve intercept), Max head (shut-off head, head at zero flow), Pump efficiency (0–1) (wire-to-water) and a Power threshold — the minimum electrical power before the motor turns.

Coupling sets how the array drives the pump:

Tank and demand

Tank capacity and Initial tank level define the reservoir buffer. Monthly water need (m³/day) is a grid of twelve cells and is the authoritative demand figure; Daily need above it is a convenience — type a value and click Fill the 12 months to set them all at once for the uniform case, then edit individual months for a seasonal profile. An optional Battery buffer (a DC buffer, with Buffer capacity and Buffer round-trip efficiency (0–1)) keeps the pump running through a cloud pass or at dawn.

Loading a pump from the catalogue

Choose from the catalogue fills the pump rows above from a library pump and records which one you used; its last entry, Add from the catalogue…, opens the catalogue's Pumps tab instead. The row appears only when the engine you are connected to publishes a pump catalogue; without it the same fields stay editable by hand. Pumps are imported from PVsyst .PMP files — see The Component Catalogue.

DC-grid traction

DC grid (rail / trolleybus) feeds the PV plant into a DC traction catenary (urban trolleybus, tram, metro or DC-electrified railway) at 600–1500 V through a DC/DC converter rather than an inverter. The defining constraint is acceptance: most traction substations use non-reversible diode rectifiers, so PV power can only be injected when a vehicle is actually drawing it. When the line is unloaded the surplus is curtailed (“DC grid not available”), unless a reversible substation lets it flow back to the AC grid. The instantaneous traction demand is therefore the acceptance budget. The simulation reports the acceptance fraction, the curtailment fraction and the traction-coverage (solar) fraction.

As with solar pumping, the consumption-rate, grid-pricing and financial cards disappear in this mode and the run books no cash flow; the note in §Solar pumping lists what that changes.

Network and converter

Network type selects the traction family: Trolleybus, Tram, Metro or Rail. It does two things — it picks the typical daily demand shape for that kind of service, and it resets Nominal voltage to the canonical catenary voltage (600 V trolleybus, 750 V tram and metro, 1500 V rail). That reset only happens when you change the network, so a project that stores a custom voltage keeps it on open.

Converter efficiency (0–1) is the flat efficiency of the DC/DC line-matching converter, and Converter threshold the minimum input power below which it stays idle (0 = no threshold).

Substation acceptance

Reversible substation is the key switch. Left off (the default), any PV power beyond the instantaneous traction demand is clipped. Ticked, the surplus flows back to the AC grid, up to Feed-back limit (0 = unlimited; the row has no effect on a non-reversible substation).

Traction demand

Peak traffic power scales the normalised daily profile of the chosen network. Leave it at 0 (an empty box shows the hint “auto (≈ PV peak)”) and the demand peak is sized on the PV production peak, so a fresh project's demand is proportional to the array rather than a scale-blind fixed figure. Weekend factor (0–1) scales the weekend demand down for reduced service.

Pre-sizing aids

Off-grid and pumping projects both need a starting point before there is anything to simulate. The Preliminary pre-sizing page (/presize) provides one: a quick monthly-resolution estimator that runs on bundled climatology, so it needs neither a weather fetch nor an open project. It has no entry in the left rail. When either the Off-grid system or the Solar pumping block is on screen, a Pre-sizing aids (off-grid / pumping)… link on the Systems card takes you there; you can also bookmark the address.

Mode switches the whole page between Off-grid (standalone) and Solar pumping. Both share the site block (Latitude, Longitude, Tilt, Azimuth, Tracker (PVGIS code 0–5), System loss) — the coordinates are pre-filled from the working project when one is open — and the Climate selector:

Grid-availability schedules

Two of the loss-cascade rows on the Production tab (/production, Running a Simulation & the Production Tab) are not flat percentages but time-varying schedules, and they are edited in the Schedules & grid section of the Configuration card beside the cascade (Schedules & grid). Click Schedule… on the Availability or Grid unavailability row of the cascade and the page scrolls to that section with the right tab already selected. A badge on the cascade row — “scheduled (N h)” or “schedule” — tells you at a glance that the flat percentage is being overridden. (The card's third tab, Grid / reactive power, is a reactive-power setpoint rather than an availability schedule and is covered with the rest of the grid connection in Electrical Design & the One-Line Diagram and Running a Simulation & the Production Tab.)

Availability — dated events & recurring profile

The Availability outage schedule tab models planned outages — “swap the inverter in 2036” — and seasonal unavailability. It has two halves that combine via a maximum: a one-shot event does not add to the seasonal profile; whichever is larger for a given hour wins.

Dated events are a table of one-off windows, one row each, with a Label, a Start (local) and End (local) date-time, and a Fraction offline between 0 and 1. Add event appends a row seeded at “whole plant offline”; the × button at the end of a row removes it. With no rows, the tab says No dated outage events.

Recurring profile (month × hour) is a 12 × 24 grid of unavailability fractions that applies every year. The full 288-cell grid is heavy, so it is folded away behind Edit grid / Hide grid. Above it, type a value and click Apply to all to fill every cell at once, or Clear to empty them — which covers the common constant-unavailability case without ever opening the grid.

Grid unavailability — weak grid

The Grid-unavailability schedule tab synthesises weak-grid outages that curtail production. Synthesis mode chooses how:

In Stochastic synthesis the outage series is drawn from a MTBF (mean interval) between events, a Mean duration (exponential law, so long outages are possible but rare), a Recovery delay added to the tail of every event for inverter resynchronisation, an Impacted fraction of the plant, and an RNG seed (0 = deterministic). At 0 the seed is derived from the project identifier and the simulation year, so the series is reproducible project by project; a fixed non-zero value locks the same series across projects and sessions.

Each of these five fields shows the engine's own default in grey when you leave it empty — an empty field means “use that value”, not “use zero”.

When the battery's Dispatch strategy is Weak-grid backup and a run has landed, a second badge on the tab — “+ N kWh battery” — reports the energy the battery rescued during outages. It appears in the Sankey waterfall as a negative loss just after Grid unavailability; see Charts, Heatmaps & Tables.

Injection (export) cap

The export limit is a rule of the grid connection, so it lives on the GRID PRICING card of the Project tab rather than in the schedules card. It is not a price: it applies whatever the card's Granularity, None included. The row shows a one-line summary of the current state — No cap, Uniform cap N kW, or Custom schedule (N capped cells) — and an Edit schedule button that unfolds the editor.

Inside is a 12-month × 24-hour grid of injection caps in kW. Uncapped cells read -1; any blank or negative cell means no cap, and 0 forbids any export in that slot. Type a value in Apply to all cells and click Apply to fill the whole grid, or Clear all to remove every cap. The cap is applied per 5-minute slot and energy above it is curtailed — the usual way to model a grid-code export limit that varies by season or time of day.

Field reference

System type (SIMULATION card)

BATTERY card

Electric-vehicle source

Reached from the Consumption tab: add a source with + and set Type to Electric vehicle.

Off-grid system (Systems card)

Backup generator (weak grid) — Systems card

Visible only in Self-consumption + storage with the Weak-grid backup dispatch strategy.

Solar pumping (Systems card)

DC traction grid (Systems card)

Pre-size page

Availability outage schedule (Production tab)

Grid-unavailability schedule (Production tab)

Injection cap (GRID PRICING card)