Light Tower Cycling Simulator
Sodium, heated lithium, and lead-acid through a full day and night in the cold
A datasheet makes this look simple. It isn't. How a battery behaves in the cold comes down to when
its heater runs, whether the generator has to run alongside it, and how those two things compound
across hundreds of cycles a year. Drag the temperature slider and watch what changes — the
story shifts completely between 40°F and 40 below. Everything updates live, and the detail behind
each number is one click away whenever you want it.
What's being compared
One 3.648 kWh Uptime Sodium battery against industry-standard lithium and lead-acid, compared
by chemistry rather than by any specific brand.
To keep it fair on usable energy, lithium is sized to the same 3.648 kWh and discharges to the same 90%. Lead-acid needs 6.566 kWh — 1.8 times as much — because discharging it past 50% sharply accelerates plate degradation. All three end up delivering the same 3.28 kWh per cycle.
The generator starts at 10% state of charge for sodium and lithium, 50% for lead-acid.
To keep it fair on usable energy, lithium is sized to the same 3.648 kWh and discharges to the same 90%. Lead-acid needs 6.566 kWh — 1.8 times as much — because discharging it past 50% sharply accelerates plate degradation. All three end up delivering the same 3.28 kWh per cycle.
The generator starts at 10% state of charge for sodium and lithium, 50% for lead-acid.
How this works & what's assumed — how the three work differently in the cold
Heater strategy, plainly:
· Sodium — heater OFF the whole discharge, draws from stored battery only for a brief pre-heat window right before the generator starts. It calculates warm-up time from ambient temp, load, and its own heat rate, then works backward so the heater switches on at exactly the right moment to reach 40°F just as SOC hits the 10% target — timed to line up with the generator start, not the other way around.
· Lithium — heater only ever runs with a charging source present; it never draws from stored capacity. No genset, no heater — the pack just cools passively toward ambient in between. Below 0°F ambient, that passive cooling would carry it through toward the -4°F hard shutdown floor, so the generator has to short-cycle throughout discharge — start, run the heater, stop — anticipating the drop rather than waiting for it. Between 0°F and 40°F, the pack never reaches the trigger on its own, so only one heat-up burst is needed at the very end, before real charging. Every burst is a no-charge inefficiency (genset on, ~0 delivered to the battery) but it keeps the pack online, and — by design — it doesn't recharge the battery, so the load still uses up real capacity between bursts rather than staying artificially topped off.
· Lead-acid — no heater at all. Cold just derates its charge acceptance and usable capacity directly; there's nothing to warm up.
· Sodium — heater OFF the whole discharge, draws from stored battery only for a brief pre-heat window right before the generator starts. It calculates warm-up time from ambient temp, load, and its own heat rate, then works backward so the heater switches on at exactly the right moment to reach 40°F just as SOC hits the 10% target — timed to line up with the generator start, not the other way around.
· Lithium — heater only ever runs with a charging source present; it never draws from stored capacity. No genset, no heater — the pack just cools passively toward ambient in between. Below 0°F ambient, that passive cooling would carry it through toward the -4°F hard shutdown floor, so the generator has to short-cycle throughout discharge — start, run the heater, stop — anticipating the drop rather than waiting for it. Between 0°F and 40°F, the pack never reaches the trigger on its own, so only one heat-up burst is needed at the very end, before real charging. Every burst is a no-charge inefficiency (genset on, ~0 delivered to the battery) but it keeps the pack online, and — by design — it doesn't recharge the battery, so the load still uses up real capacity between bursts rather than staying artificially topped off.
· Lead-acid — no heater at all. Cold just derates its charge acceptance and usable capacity directly; there's nothing to warm up.
Sodium (hybrid, §8.2): heater OFF except a brief pre-heat window before genset start
(drawn from stored charge, not genset fuel). Charges at a fixed 80A/28.4V — same rate as
lithium, so recharge time isn't an artifact of which genset it's paired with. Genset modeled as
a Hatz 1B30 DC unit (3,000W), tuned for peak efficiency (illustrative BSFC 0.08 gal/kWh at
80–100% load).
Lithium (heated, generator-sourced heat): heater only
ever runs with a charging source present — it never self-consumes stored capacity (corrected
from an earlier version of this model). No genset, no heater; the pack just cools passively
toward ambient in between. Hard shutdown floor at -4°F, unconfirmed. Below 0°F ambient, passive
cooling would carry the pack through that floor, so the generator has to short-cycle
throughout discharge — start, run the heater (and carry the load) back up to 40°F, stop —
anticipating the drop rather than reacting to it. These bursts don't recharge the battery, so
real capacity still gets used by the load between them (the alternative — recharging on every
short cycle — would keep SOC topped up so shallowly the battery's capacity barely gets used;
not modeled by default). Between 0°F and 40°F ambient, the pack never reaches the restart
trigger on its own, so only a single heat-up burst is needed at the very end, before real
charging — same shape as sodium's pre-heat window, just genset-sourced instead of
battery-sourced. Heater modeled at a weaker 150W/unit ("little heater," illustrative). Bank
sized on equal nameplate Wh vs. sodium — the project's equal-energy-basis question is still
open. Charges at the same fixed 80A/28.4V as sodium — no chemistry-based charge-rate
advantage. Genset modeled as a Kubota D902-class 3-cylinder diesel unit (6,000W, illustrative
BSFC 0.12 gal/kWh, or smaller); these heater bursts run at near-zero delivered power, so the
light-load efficiency penalty makes them disproportionately costly and they extend total
genset runtime every cycle. Rated at 4,000 cycles (user-provided); 80% DoD basis assumed, not
confirmed. Passive cooling rate (Newton's-law constant) is illustrative and unconfirmed — no
real thermal data exists for this pack.
Lead-acid (conventional): no active heater — floor is charge acceptance and discharge
capacity, not cell protection. Bank sized at 6,566Wh nameplate (1.8× sodium/lithium's 3,648Wh)
on an equal-usable-energy basis: both deliver ~3,283Wh per cycle once sodium/lithium's 90% DoD
and lead-acid's 50% DoD are applied. DoD floor fixed at 50% — never adjustable — because
discharging past that sharply accelerates plate degradation, not a safety cutoff. Charges at
the same fixed 80A/28.4V baseline as
sodium/lithium — derated only by the charge-acceptance curve below, not by genset size. Two
separate cold-weather derates apply, both illustrative and unconfirmed (no manufacturer curve
exists in the project): charge-acceptance rate (how much of that 80A it can actually accept,
down to 15% at -38°F) and discharge capacity (how much usable energy is actually available,
down to 50% of nameplate at -38°F) — these are different mechanisms and both get worse
independently as it gets colder. Same Kubota D902-class 3-cylinder genset assumption as
lithium.
Non-hybrid benchmark (no battery): genset runs directly whenever lights are on, off
otherwise. No cycling logic, no pre-heat, no charge-acceptance question — it's just diesel
powering LEDs. Same Kubota D902-class 3-cylinder genset assumption (6,000W nameplate, or
smaller), illustrative BSFC 0.12 gal/kWh at peak efficiency.
Fuel model (all four configs): fuel is computed from actual power delivered by the
genset in each phase, not nameplate capacity, with a shared efficiency-penalty curve — every
genset is assumed tuned for peak efficiency at 80–100% of its own rated load, and burns
progressively worse per kWh delivered below that. This replaces an earlier version of this
tool that priced hybrid fuel off nameplate output regardless of actual load, which understated
the fuel cost of a genset carrying only a small charge current. Curve is illustrative, not a
manufacturer efficiency map.
Day/night load: night hours run full LED load; day hours run a fixed 25W controller/idle draw
(assumption, battery configs only — non-hybrid has no day-hour draw). This tool averages load
across a full day rather than time-stepping hour by hour.
Full per-config data table — every number this tool computes
| Per-cycle / annual (modeled) | Sodium | Lithium (heated) | Lead-acid | Non-hybrid (no battery) |
|---|---|---|---|---|
| DoD floor per cycle (%) | – | – | 50 (fixed) | n/a — no battery |
| Discharge capacity at this ambient (% of nameplate) | n/a | n/a | – | n/a |
| Idle/pre-heat delay before charge (hrs) — genset OFF (sodium), genset ON/idling (lithium) | – | – | 0 (no heater) | n/a |
| Discharge phase (hrs) | – | – | – | n/a |
| Recharge phase only (hrs) | – | – | – | – |
| Total genset runtime per cycle (hrs) — idle + recharge, where genset is actually on | – | – | – | – |
| Total cycle (hrs) | – | – | – | 24 (fixed) |
| Cycles / year | – | – | – | – |
| Generator starts per night — at this ambient/load | – | – | – | 1 (continuous run) |
| Generator starts / year | – | – | – | – |
| Short cycles per discharge — heater-only, no charge, below 0°F ambient | n/a — pre-heats from stored charge, no cycling | – | n/a — no heater | n/a |
| Generator hours / year | – | – | – | – |
| Genset load fraction / efficiency penalty | – | – | – | – |
| Fuel, gal/year (modeled) | – | – | – | – |
| Modeled life at this duty (yrs) | – | – | – | n/a — no battery |
Fuel across the temperature range
Modeled fuel burn per year (gal, vertical axis) as ambient temperature drops. Watch where the lines separate.
How to read this chart
Fuel is calculated from the power the generator actually delivers, not its nameplate
rating, and every configuration pays the same efficiency penalty when its engine runs
underloaded. Sodium runs a small Hatz 1B30; the other three run a larger Kubota D902-class
unit. That engine pairing drives most of the separation you see, so it's worth confirming
against real hardware before quoting these figures.
One full cycle, second by second
Fuel burned across a single cycle. The red segment is the warm-up window before charging can start — and it behaves completely differently for each battery.
Why the red segments look so different
Sodium's red segment is flat. It warms itself from
stored battery charge with the generator switched off, so the warm-up costs no fuel at
all.
Lithium's red segment climbs. Its heater can only run when a charging source is present, so the generator has to be running the whole time — burning fuel at almost no useful output while it waits for the cells to warm up.
That difference is the single clearest thing this simulation shows. Everything else — fuel totals, generator starts, runtime — follows from it.
Lines are drawn straight between phase boundaries rather than fully time-stepped. Each line ends when that configuration's own cycle ends, so a shorter line means the generator restarts sooner, not that it does less work.
Lithium's red segment climbs. Its heater can only run when a charging source is present, so the generator has to be running the whole time — burning fuel at almost no useful output while it waits for the cells to warm up.
That difference is the single clearest thing this simulation shows. Everything else — fuel totals, generator starts, runtime — follows from it.
Lines are drawn straight between phase boundaries rather than fully time-stepped. Each line ends when that configuration's own cycle ends, so a shorter line means the generator restarts sooner, not that it does less work.
Fuel per cycle
How much less fuel sodium burns than each alternative, at the current settings.
–%
less fuel than heated lithium
–% from the battery and heater ·
–% from the engine
–%
less fuel than lead-acid
–% from the battery ·
–% from the engine
–%
less fuel than running the generator non-stop
No battery to compare — this is the whole hybrid saving.
What the two smaller numbers mean — two reasons sodium wins, not one
Sodium wins here for two separate reasons, and it's worth knowing which is which. Swap
lithium onto sodium's own Hatz generator and hold the chemistry fixed: whatever gap closes
from that swap alone is the engine's share. Whatever gap is
still left once both run the same engine is the battery and
heater's share.
Above freezing that split lands almost entirely on the engine — no heater is running, so chemistry barely matters. As it gets colder, the battery share grows, because that's when lithium's heater starts forcing the generator to run. Move the temperature slider and watch the two numbers trade places.
Cards below a 5% difference are hidden — at those settings the two options are close enough that the gap isn't meaningful. Every figure here depends on generator fuel curves and heater wattages that aren't confirmed yet, so treat them as directional.
Above freezing that split lands almost entirely on the engine — no heater is running, so chemistry barely matters. As it gets colder, the battery share grows, because that's when lithium's heater starts forcing the generator to run. Move the temperature slider and watch the two numbers trade places.
Cards below a 5% difference are hidden — at those settings the two options are close enough that the gap isn't meaningful. Every figure here depends on generator fuel curves and heater wattages that aren't confirmed yet, so treat them as directional.
Generator activity
How much less the generator has to run and restart.
Sodium vs LITHIUM
–%
fewer generator starts/yr, sodium vs. lithium
–%
more hours between starts, sodium vs. lithium
–%
shorter runtime per start, sodium vs. lithium
Sodium vs LEAD-ACID
–%
fewer generator starts/yr, sodium vs. lead-acid
–%
more hours between starts, sodium vs. lead-acid
–%
shorter runtime per start, sodium vs. lead-acid
Sodium vs NON-HYBRID
–%
fewer generator starts/yr, sodium vs. non-hybrid
–%
more hours between starts, sodium vs. non-hybrid
–%
less generator runtime per night than non-hybrid running continuously
What these three numbers mean
Fewer starts and more hours between starts are two views of the same thing — how long a
full cycle lasts — so they move together. Runtime per start is separate: it's how long the
generator actually runs once it fires up, including any time spent warming the battery
before charging can begin. Cards below a 5% difference are hidden.
Per-cycle fuel breakdown — where the fuel goes
| Config | Generator | Fuel / cycle (gal) |
|---|---|---|
| Sodium (pre-heats from stored battery, genset off) | Hatz 1B30, 3,000W | – |
| Lithium, as configured (idle-to-heat + recharge) | Kubota D902-class, 6,000W | – |
| ↳ of which, idle-to-heat alone ("idle tax") | – | |
| Lead-acid | Kubota D902-class, 6,000W | – |
| Non-hybrid (24h clock, no battery) | Kubota D902-class, 6,000W | – |
The idle tax is –% of lithium's total per-cycle fuel — that
part is heating-related, and disappears above 40°F. But sodium is also paired with a smaller,
more efficient generator (Hatz 1B30, 0.08 gal/kWh peak) than the other three configs (Kubota
D902-class, 0.12 gal/kWh peak). That gap doesn't go away in warm weather — it's why sodium
still burns less fuel even when nobody needs a heater. Both effects are real advantages of this
design, but they're two different mechanisms, and the generator-pairing assumption is the one
most worth confirming against real hardware before this number goes external.
Pre-heat from battery vs. idle the generator — genset off vs. genset idling
Pre-heating from stored battery charge always costs 0 direct generator fuel — the genset stays off. The cells below show what that same heater/duration would cost if the generator had to idle and supply the heater instead, on each engine. That's the direct gallons advantage of pre-heating, isolated from which battery chemistry or genset you're comparing.
| Heater / heat-up duration this cycle | Duration (hrs) | Idle cost on Hatz 1B30 (gal) | Idle cost on Kubota D902-class (gal) |
|---|---|---|---|
| Sodium's heater (350W/unit) | – | – | – |
| Lithium's heater (150W/unit, weaker) | – | – | – |
Every cell in this table is the gallons pre-heating avoids, for that heater/duration/engine
combination, at the current ambient and load. The Kubota column is worse per gallon at low
load (light-load penalty) and has a higher base BSFC than the Hatz — so the same idle window
costs more on the Kubota regardless of which battery it's paired with. This ignores
wet-stacking risk entirely: idling a cold diesel engine has real engine-wear and fuel-dilution
costs beyond what's counted here, which is a second, separate reason (SPEC_MASTER §8.2 intent)
to avoid it, not counted in these gallon figures.
Light Tower Cycling Simulator
Sodium, heated lithium, and lead-acid through a full day and night in the cold
A datasheet makes this look simple. It isn't. How a battery behaves in the cold comes down to when
its heater runs, whether the generator has to run alongside it, and how those two things compound
across hundreds of cycles a year. Drag the temperature slider and watch what changes — the
story shifts completely between 40°F and 40 below. Everything updates live, and the detail behind
each number is one click away whenever you want it.
What's being compared
One 3.648 kWh Uptime Sodium battery against industry-standard lithium and lead-acid, compared
by chemistry rather than by any specific brand.
To keep it fair on usable energy, lithium is sized to the same 3.648 kWh and discharges to the same 90%. Lead-acid needs 6.566 kWh — 1.8 times as much — because discharging it past 50% sharply accelerates plate degradation. All three end up delivering the same 3.28 kWh per cycle.
The generator starts at 10% state of charge for sodium and lithium, 50% for lead-acid.
To keep it fair on usable energy, lithium is sized to the same 3.648 kWh and discharges to the same 90%. Lead-acid needs 6.566 kWh — 1.8 times as much — because discharging it past 50% sharply accelerates plate degradation. All three end up delivering the same 3.28 kWh per cycle.
The generator starts at 10% state of charge for sodium and lithium, 50% for lead-acid.
How this works & what's assumed — how the three work differently in the cold
Heater strategy, plainly:
· Sodium — heater OFF the whole discharge, draws from stored battery only for a brief pre-heat window right before the generator starts. It calculates warm-up time from ambient temp, load, and its own heat rate, then works backward so the heater switches on at exactly the right moment to reach 40°F just as SOC hits the 10% target — timed to line up with the generator start, not the other way around.
· Lithium — heater only ever runs with a charging source present; it never draws from stored capacity. No genset, no heater — the pack just cools passively toward ambient in between. Below 0°F ambient, that passive cooling would carry it through toward the -4°F hard shutdown floor, so the generator has to short-cycle throughout discharge — start, run the heater, stop — anticipating the drop rather than waiting for it. Between 0°F and 40°F, the pack never reaches the trigger on its own, so only one heat-up burst is needed at the very end, before real charging. Every burst is a no-charge inefficiency (genset on, ~0 delivered to the battery) but it keeps the pack online, and — by design — it doesn't recharge the battery, so the load still uses up real capacity between bursts rather than staying artificially topped off.
· Lead-acid — no heater at all. Cold just derates its charge acceptance and usable capacity directly; there's nothing to warm up.
· Sodium — heater OFF the whole discharge, draws from stored battery only for a brief pre-heat window right before the generator starts. It calculates warm-up time from ambient temp, load, and its own heat rate, then works backward so the heater switches on at exactly the right moment to reach 40°F just as SOC hits the 10% target — timed to line up with the generator start, not the other way around.
· Lithium — heater only ever runs with a charging source present; it never draws from stored capacity. No genset, no heater — the pack just cools passively toward ambient in between. Below 0°F ambient, that passive cooling would carry it through toward the -4°F hard shutdown floor, so the generator has to short-cycle throughout discharge — start, run the heater, stop — anticipating the drop rather than waiting for it. Between 0°F and 40°F, the pack never reaches the trigger on its own, so only one heat-up burst is needed at the very end, before real charging. Every burst is a no-charge inefficiency (genset on, ~0 delivered to the battery) but it keeps the pack online, and — by design — it doesn't recharge the battery, so the load still uses up real capacity between bursts rather than staying artificially topped off.
· Lead-acid — no heater at all. Cold just derates its charge acceptance and usable capacity directly; there's nothing to warm up.
Sodium (hybrid, §8.2): heater OFF except a brief pre-heat window before genset start
(drawn from stored charge, not genset fuel). Charges at a fixed 80A/28.4V — same rate as
lithium, so recharge time isn't an artifact of which genset it's paired with. Genset modeled as
a Hatz 1B30 DC unit (3,000W), tuned for peak efficiency (illustrative BSFC 0.08 gal/kWh at
80–100% load).
Lithium (heated, generator-sourced heat): heater only
ever runs with a charging source present — it never self-consumes stored capacity (corrected
from an earlier version of this model). No genset, no heater; the pack just cools passively
toward ambient in between. Hard shutdown floor at -4°F, unconfirmed. Below 0°F ambient, passive
cooling would carry the pack through that floor, so the generator has to short-cycle
throughout discharge — start, run the heater (and carry the load) back up to 40°F, stop —
anticipating the drop rather than reacting to it. These bursts don't recharge the battery, so
real capacity still gets used by the load between them (the alternative — recharging on every
short cycle — would keep SOC topped up so shallowly the battery's capacity barely gets used;
not modeled by default). Between 0°F and 40°F ambient, the pack never reaches the restart
trigger on its own, so only a single heat-up burst is needed at the very end, before real
charging — same shape as sodium's pre-heat window, just genset-sourced instead of
battery-sourced. Heater modeled at a weaker 150W/unit ("little heater," illustrative). Bank
sized on equal nameplate Wh vs. sodium — the project's equal-energy-basis question is still
open. Charges at the same fixed 80A/28.4V as sodium — no chemistry-based charge-rate
advantage. Genset modeled as a Kubota D902-class 3-cylinder diesel unit (6,000W, illustrative
BSFC 0.12 gal/kWh, or smaller); these heater bursts run at near-zero delivered power, so the
light-load efficiency penalty makes them disproportionately costly and they extend total
genset runtime every cycle. Rated at 4,000 cycles (user-provided); 80% DoD basis assumed, not
confirmed. Passive cooling rate (Newton's-law constant) is illustrative and unconfirmed — no
real thermal data exists for this pack.
Lead-acid (conventional): no active heater — floor is charge acceptance and discharge
capacity, not cell protection. Bank sized at 6,566Wh nameplate (1.8× sodium/lithium's 3,648Wh)
on an equal-usable-energy basis: both deliver ~3,283Wh per cycle once sodium/lithium's 90% DoD
and lead-acid's 50% DoD are applied. DoD floor fixed at 50% — never adjustable — because
discharging past that sharply accelerates plate degradation, not a safety cutoff. Charges at
the same fixed 80A/28.4V baseline as
sodium/lithium — derated only by the charge-acceptance curve below, not by genset size. Two
separate cold-weather derates apply, both illustrative and unconfirmed (no manufacturer curve
exists in the project): charge-acceptance rate (how much of that 80A it can actually accept,
down to 15% at -38°F) and discharge capacity (how much usable energy is actually available,
down to 50% of nameplate at -38°F) — these are different mechanisms and both get worse
independently as it gets colder. Same Kubota D902-class 3-cylinder genset assumption as
lithium.
Non-hybrid benchmark (no battery): genset runs directly whenever lights are on, off
otherwise. No cycling logic, no pre-heat, no charge-acceptance question — it's just diesel
powering LEDs. Same Kubota D902-class 3-cylinder genset assumption (6,000W nameplate, or
smaller), illustrative BSFC 0.12 gal/kWh at peak efficiency.
Fuel model (all four configs): fuel is computed from actual power delivered by the
genset in each phase, not nameplate capacity, with a shared efficiency-penalty curve — every
genset is assumed tuned for peak efficiency at 80–100% of its own rated load, and burns
progressively worse per kWh delivered below that. This replaces an earlier version of this
tool that priced hybrid fuel off nameplate output regardless of actual load, which understated
the fuel cost of a genset carrying only a small charge current. Curve is illustrative, not a
manufacturer efficiency map.
Day/night load: night hours run full LED load; day hours run a fixed 25W controller/idle draw
(assumption, battery configs only — non-hybrid has no day-hour draw). This tool averages load
across a full day rather than time-stepping hour by hour.
Full per-config data table — every number this tool computes
| Per-cycle / annual (modeled) | Sodium | Lithium (heated) | Lead-acid | Non-hybrid (no battery) |
|---|---|---|---|---|
| DoD floor per cycle (%) | – | – | 50 (fixed) | n/a — no battery |
| Discharge capacity at this ambient (% of nameplate) | n/a | n/a | – | n/a |
| Idle/pre-heat delay before charge (hrs) — genset OFF (sodium), genset ON/idling (lithium) | – | – | 0 (no heater) | n/a |
| Discharge phase (hrs) | – | – | – | n/a |
| Recharge phase only (hrs) | – | – | – | – |
| Total genset runtime per cycle (hrs) — idle + recharge, where genset is actually on | – | – | – | – |
| Total cycle (hrs) | – | – | – | 24 (fixed) |
| Cycles / year | – | – | – | – |
| Generator starts per night — at this ambient/load | – | – | – | 1 (continuous run) |
| Generator starts / year | – | – | – | – |
| Short cycles per discharge — heater-only, no charge, below 0°F ambient | n/a — pre-heats from stored charge, no cycling | – | n/a — no heater | n/a |
| Generator hours / year | – | – | – | – |
| Genset load fraction / efficiency penalty | – | – | – | – |
| Fuel, gal/year (modeled) | – | – | – | – |
| Modeled life at this duty (yrs) | – | – | – | n/a — no battery |
Fuel across the temperature range
Modeled fuel burn per year (gal, vertical axis) as ambient temperature drops. Watch where the lines separate.
How to read this chart
Fuel is calculated from the power the generator actually delivers, not its nameplate
rating, and every configuration pays the same efficiency penalty when its engine runs
underloaded. Sodium runs a small Hatz 1B30; the other three run a larger Kubota D902-class
unit. That engine pairing drives most of the separation you see, so it's worth confirming
against real hardware before quoting these figures.
One full cycle, second by second
Fuel burned across a single cycle. The red segment is the warm-up window before charging can start — and it behaves completely differently for each battery.
Why the red segments look so different
Sodium's red segment is flat. It warms itself from
stored battery charge with the generator switched off, so the warm-up costs no fuel at
all.
Lithium's red segment climbs. Its heater can only run when a charging source is present, so the generator has to be running the whole time — burning fuel at almost no useful output while it waits for the cells to warm up.
That difference is the single clearest thing this simulation shows. Everything else — fuel totals, generator starts, runtime — follows from it.
Lines are drawn straight between phase boundaries rather than fully time-stepped. Each line ends when that configuration's own cycle ends, so a shorter line means the generator restarts sooner, not that it does less work.
Lithium's red segment climbs. Its heater can only run when a charging source is present, so the generator has to be running the whole time — burning fuel at almost no useful output while it waits for the cells to warm up.
That difference is the single clearest thing this simulation shows. Everything else — fuel totals, generator starts, runtime — follows from it.
Lines are drawn straight between phase boundaries rather than fully time-stepped. Each line ends when that configuration's own cycle ends, so a shorter line means the generator restarts sooner, not that it does less work.
Fuel per cycle
How much less fuel sodium burns than each alternative, at the current settings.
–%
less fuel than heated lithium
–% from the battery and heater ·
–% from the engine
–%
less fuel than lead-acid
–% from the battery ·
–% from the engine
–%
less fuel than running the generator non-stop
No battery to compare — this is the whole hybrid saving.
What the two smaller numbers mean — two reasons sodium wins, not one
Sodium wins here for two separate reasons, and it's worth knowing which is which. Swap
lithium onto sodium's own Hatz generator and hold the chemistry fixed: whatever gap closes
from that swap alone is the engine's share. Whatever gap is
still left once both run the same engine is the battery and
heater's share.
Above freezing that split lands almost entirely on the engine — no heater is running, so chemistry barely matters. As it gets colder, the battery share grows, because that's when lithium's heater starts forcing the generator to run. Move the temperature slider and watch the two numbers trade places.
Cards below a 5% difference are hidden — at those settings the two options are close enough that the gap isn't meaningful. Every figure here depends on generator fuel curves and heater wattages that aren't confirmed yet, so treat them as directional.
Above freezing that split lands almost entirely on the engine — no heater is running, so chemistry barely matters. As it gets colder, the battery share grows, because that's when lithium's heater starts forcing the generator to run. Move the temperature slider and watch the two numbers trade places.
Cards below a 5% difference are hidden — at those settings the two options are close enough that the gap isn't meaningful. Every figure here depends on generator fuel curves and heater wattages that aren't confirmed yet, so treat them as directional.
Generator activity
How much less the generator has to run and restart.
Sodium vs LITHIUM
–%
fewer generator starts/yr, sodium vs. lithium
–%
more hours between starts, sodium vs. lithium
–%
shorter runtime per start, sodium vs. lithium
Sodium vs LEAD-ACID
–%
fewer generator starts/yr, sodium vs. lead-acid
–%
more hours between starts, sodium vs. lead-acid
–%
shorter runtime per start, sodium vs. lead-acid
Sodium vs NON-HYBRID
–%
fewer generator starts/yr, sodium vs. non-hybrid
–%
more hours between starts, sodium vs. non-hybrid
–%
less generator runtime per night than non-hybrid running continuously
What these three numbers mean
Fewer starts and more hours between starts are two views of the same thing — how long a
full cycle lasts — so they move together. Runtime per start is separate: it's how long the
generator actually runs once it fires up, including any time spent warming the battery
before charging can begin. Cards below a 5% difference are hidden.
Per-cycle fuel breakdown — where the fuel goes
| Config | Generator | Fuel / cycle (gal) |
|---|---|---|
| Sodium (pre-heats from stored battery, genset off) | Hatz 1B30, 3,000W | – |
| Lithium, as configured (idle-to-heat + recharge) | Kubota D902-class, 6,000W | – |
| ↳ of which, idle-to-heat alone ("idle tax") | – | |
| Lead-acid | Kubota D902-class, 6,000W | – |
| Non-hybrid (24h clock, no battery) | Kubota D902-class, 6,000W | – |
The idle tax is –% of lithium's total per-cycle fuel — that
part is heating-related, and disappears above 40°F. But sodium is also paired with a smaller,
more efficient generator (Hatz 1B30, 0.08 gal/kWh peak) than the other three configs (Kubota
D902-class, 0.12 gal/kWh peak). That gap doesn't go away in warm weather — it's why sodium
still burns less fuel even when nobody needs a heater. Both effects are real advantages of this
design, but they're two different mechanisms, and the generator-pairing assumption is the one
most worth confirming against real hardware before this number goes external.
Pre-heat from battery vs. idle the generator — genset off vs. genset idling
Pre-heating from stored battery charge always costs 0 direct generator fuel — the genset stays off. The cells below show what that same heater/duration would cost if the generator had to idle and supply the heater instead, on each engine. That's the direct gallons advantage of pre-heating, isolated from which battery chemistry or genset you're comparing.
| Heater / heat-up duration this cycle | Duration (hrs) | Idle cost on Hatz 1B30 (gal) | Idle cost on Kubota D902-class (gal) |
|---|---|---|---|
| Sodium's heater (350W/unit) | – | – | – |
| Lithium's heater (150W/unit, weaker) | – | – | – |
Every cell in this table is the gallons pre-heating avoids, for that heater/duration/engine
combination, at the current ambient and load. The Kubota column is worse per gallon at low
load (light-load penalty) and has a higher base BSFC than the Hatz — so the same idle window
costs more on the Kubota regardless of which battery it's paired with. This ignores
wet-stacking risk entirely: idling a cold diesel engine has real engine-wear and fuel-dilution
costs beyond what's counted here, which is a second, separate reason (SPEC_MASTER §8.2 intent)
to avoid it, not counted in these gallon figures.