About the Author
dosanko_tousan. 50 years old, stay-at-home father, non-engineer, born in Iwamizawa. Independent AI alignment researcher (GLG Network, Zenodo DOI: 10.5281/zenodo.18691357). This series is not a side project from AI research. It is written as repayment to Hokkaido.
Introduction: From Iwamizawa
In the 1990s, I watched the coal industry die in Iwamizawa.
Mines closed. The town grew quiet. What would Hokkaido's energy become? — I've carried that question ever since.
One answer is this series.
Hokkaido's renewable energy potential is the highest in Japan. Solar #1 nationally, wind #1 nationally. But that electricity is being thrown away through output curtailment because it can't be used up.
Meanwhile, EV adoption rate is dead last nationally. 4.3 vehicles per 10,000 people.
Surplus electricity and cars that should be running are not connected — blocked by a wall called "policy."
This series breaks that wall with physics and numbers. That's the job.
0. Hokkaido EV Special Zone Concept — Overview of the "Five Arrows"
Conclusions first. Evidence is built in subsequent chapters.
Arrow ① Legislating the "Right to Charge"
The current system allowing condominium management associations to veto EV charger installation is the biggest barrier. Norway banned this by law in 2017. Referencing Norway's Right-to-Plug approach as a regulatory reform exception under the Hokkaido GX Special Zone: adopt a system design where "if reasonable technical requirements and a cost-sharing plan are submitted, the management association cannot refuse without objective reasons." Implementation feasibility is ensured by conditioning on cost sharing, maintenance responsibility, and fire safety design.
Arrow ② Creating a "Cold Climate Coefficient" Subsidy
The national CEV subsidy is uniform across Japan. However, as detailed below, battery range decreases 20–40% during Hokkaido winters. The additional costs for cold-climate specifications, battery preconditioning systems, and 4WD are orders of magnitude different from Honshu. Establish a special zone-specific supplement that multiplies the base subsidy by a cold climate coefficient (×1.5–2.0).
Arrow ③ V2H Disaster Prevention Infrastructure Subsidy
September 6, 2018, Hokkaido Eastern Iburi Earthquake. Japan's first blackout — 2.95 million households without power across Hokkaido, 45 hours until restoration. If an earthquake of equal magnitude hits in winter, loss of heating becomes a life-or-death issue. A single EV (60kWh) can supply 2–3 days of electricity for a typical household via V2H. Establish a special "disaster prevention purpose subsidy" category for bundled V2H equipment and EV purchases.
Arrow ④ Eliminating "Charging Dead Zones"
Of Hokkaido's 127 roadside stations (michi-no-eki), only 51 (40%) have EV charging. In a land where getting stranded in winter is life-threatening, this is unacceptable. Mandate maximum 50km charging intervals on trunk roads and installation of rapid chargers (50kW+, with snow-melting capability) at all roadside stations.
Arrow ⑤ Direct Utilization of Surplus Renewable Energy for EV Charging
Hokkaido's solar connection capacity is 2.31 GW, exceeding the connectable capacity of 1.17 GW by over 1 GW. During low-demand periods like Golden Week, renewable output curtailment is implemented. Electricity is surplus and being thrown away. Implement a DR (demand response) mechanism that offers rebates (post-hoc reimbursement) or reduced kWh rates for EV charging during curtailment hours. This simultaneously resolves both "EVs are expensive to charge" and "surplus renewables are being wasted" through a coherent policy design. An excellent fit for the GX Special Zone's "advanced technology demonstration and implementation" framework.
Why this isn't "fantasy"
In June 2024, all of Hokkaido was designated as a National Strategic Special Zone for "GX Finance and Asset Management." No need to apply from zero. Incorporate "EV operations" into the existing GX Special Zone. Under the National Strategic Special Zone system, anyone — municipalities, companies, or individuals — can propose regulatory reform ideas, with submissions accepted on a rolling basis.
1. Why Lithium-Ion Batteries Are Weak Against Cold
1.1 The Arrhenius Equation — Exponential Temperature Dependence
The operating principle of lithium-ion batteries is simple. Lithium ions stored in the anode (graphite) swim through the electrolyte to reach the cathode. The rate-limiting step of battery performance is "how fast the ions swim" — ionic conductivity.
The temperature dependence of ionic conductivity is described by the Arrhenius equation:
$$\sigma(T) = \sigma_0 \exp\left(-\frac{E_a}{k_B T}\right)$$
$E_a$ is activation energy, $k_B$ is the Boltzmann constant ($1.381 \times 10^{-23}$ J/K), and $T$ is absolute temperature (K).
Because temperature sits in the exponent, the impact of temperature decrease is not linear but exponential.
For typical lithium-ion battery electrolyte (EC/DMC system, LiPF₆ 1M), activation energy is approximately 0.2–0.4 eV. Setting $E_a = 0.3 \text{ eV}$, the ionic conductivity ratio between 25°C (298K) and -20°C (253K):
$$\frac{\sigma(253)}{\sigma(298)} = \exp\left(-\frac{E_a}{k_B}\left(\frac{1}{253} - \frac{1}{298}\right)\right) = \exp(-2.08) \approx 0.125$$
About 12.6% of the 25°C value — at -20°C, ionic conductivity drops to approximately one-eighth of room temperature.
import numpy as np
def conductivity_ratio(T_celsius: float, T_ref: float = 25.0) -> float:
"""Relative conductivity to 25°C (Arrhenius equation)"""
k_B = 1.381e-23
eV_to_J = 1.602e-19
E_a_J = 0.3 * eV_to_J
T_K = T_celsius + 273.15
T_ref_K = T_ref + 273.15
return np.exp(-E_a_J / k_B * (1/T_K - 1/T_ref_K))
# Hokkaido major cities, January average temperatures
hokkaido_cities = {
"Sapporo": -3.6, "Asahikawa": -7.5, "Obihiro": -7.5,
"Kushiro": -5.9, "Wakkanai": -4.5, "Rikubetsu (coldest)": -15.0,
}
for city, temp in hokkaido_cities.items():
ratio = conductivity_ratio(temp)
print(f"{city:20s} {temp:6.1f}°C conductivity: {ratio:.3f} ({1/ratio:.1f}x slower)")
Note: Ionic conductivity reduction does not directly equal range loss. Actual loss is determined by the sum of (a) output limitation from internal resistance increase, (b) power consumption for preheating and cabin heating, and (c) BMS charging speed limitations. The Arrhenius equation is positioned as supporting data providing the basis for (a).
1.2 Electrolyte Freezing Points — The Invisible Limit Line
| Component | Freezing Point |
|---|---|
| EC (Ethylene Carbonate) | +36.4°C |
| DMC (Dimethyl Carbonate) | +4.6°C |
| EC/DMC mix (1:1) | approx. -20°C |
| PC (Propylene Carbonate) | -48.8°C |
The EC/DMC system freezing point is approximately -20°C. This perfectly coincides with Hokkaido's deep winter outdoor temperatures.
Real vehicle battery packs have thermal management systems, so the electrolyte typically doesn't freeze. However, after overnight outdoor parking, large amounts of energy are consumed for battery preconditioning (preheating) before driving is possible.
If 5kWh is used for preheating from a 60kWh battery, available driving energy is 55kWh — 8% range loss at this point alone. Combined with heater cabin consumption, 20–30% loss is fully explained by this alone.
1.3 Lithium Plating — The Hidden Danger of Cold-Weather Fast Charging
The most dangerous phenomenon in low-temperature environments is lithium plating during fast charging.
At room temperature, lithium ions during charging are orderly intercalated into the graphite anode's interlayer spaces. But at low temperatures, ion diffusion speed drops and intercalation can't keep up, causing metallic lithium to deposit on the anode surface.
Deposited lithium causes:
- Irreversible capacity loss (the battery's "stamina" permanently decreases)
- Becomes the starting point for dendrite (tree-like crystal) growth, which in worst cases can pierce the separator and lead to internal short circuit → thermal runaway
This is why the BMS drastically limits charging speed during low-temperature fast charging. Winter reports of "Tesla Supercharger should output 250kW V3 but only getting 50kW" are the result of the BMS intentionally reducing charging speed to prevent lithium plating.
Implication for Hokkaido: Winter outdoor fast charging takes 2–3x longer than Honshu. Waiting 30 minutes vs. 1 hour at -20°C outdoors — this is not a "convenience" issue but a "safety" issue. This is the physical basis for why mandating roofs, windbreaks, and snow-melting equipment at chargers (Arrow ④) is necessary.
2. NAF Test Data — The Reality Told by Real Cars
2.1 NAF El Prix 2025 (-6°C to +8°C, 24 vehicles)
The Norwegian Automobile Federation (NAF) conducts real-world EV range tests every winter.
| Rank | Vehicle | WLTP Loss |
|---|---|---|
| 1st (min loss) | BMW iX3 | -4% |
| — | Polestar 3 | -5% |
| — | Hyundai Kona Electric | approx. -9% |
| Upper loss range | Tesla Model 3 LR | approx. -24% |
| Max loss | Opel Ampera-e | -29% |
The spread of losses from -4% to -29% (7x+ difference) is the most important finding. "EVs are weak against cold" is only half correct. Precisely: "EVs with insufficient cold-weather measures are weak against cold."
2.2 NAF El Prix 2026 (-31°C, 24 vehicles) — Lowest Temperature Test in History
| Rank | Vehicle | WLTP Loss |
|---|---|---|
| Min loss | MG 6S | -29% |
| Max loss | Lucid Air / Opel Grandland | -46% |
At -31°C, all vehicles fell below WLTP values. Even the best performer, MG 6S, lost 29%.
3. Nissan Leaf e+ Real-World Data — The Reality for Hokkaido Owners
| Season | Efficiency (km/kWh) | 60kWh Equivalent Range |
|---|---|---|
| Spring–Autumn | 7.4 | 444 km |
| Winter | 5.1 | 306 km |
| Loss rate | ▲31% |
444km in spring drops to 306km in winter. 31% measured loss.
For the older Leaf with WLTP 322km, if winter range drops below 200km, even the one-way trip from Sapporo to Asahikawa (~140km) may require mid-route charging. And as noted, winter fast charging takes 2–3x longer.
This is the real reason "EVs don't sell in Hokkaido."
Not cold itself. The triple threat of cold × insufficient charging infrastructure × range anxiety hits simultaneously.
4. What Norway Proved — 97% in "The Same Cold"
4.1 30 Years of Consistency
Norway's EV adoption reached 97% of new car sales in 2025. The essence is 30 years of consistent policy.
| Year | Policy |
|---|---|
| 1990 | EV purchase tax / import tax exemption |
| 1997 | Toll road free |
| 2001 | VAT (25%) exemption |
| 2005 | Bus lane access |
| 2009 | Ferry free / corporate vehicle tax 50% reduction |
| 2017 | Legislated condominium residents' "Right to Charge" |
Policy didn't change even when governments changed. This is the essence of Norway's success.
4.2 Population Scale Comparison — Same Playing Field as Hokkaido
| Metric | Norway | Hokkaido |
|---|---|---|
| Population | ~5.5M | ~5.2M |
| Area | 385,000 km² | 83,000 km² |
| Population density | 14/km² | 63/km² |
| Public charging points | 24,000+ | Est. <1,500 |
| Chargers per 100K pop. | 436 | Est. <30 |
| Registered EVs | 800,000+ | ~2,200 |
Nearly identical population. Norway's area is 4.6x larger. Yet charging infrastructure is 15x+ greater than Hokkaido.
This is the concrete picture of the policy gap.
5. Hokkaido GX Special Zone — An Unused Weapon
5.1 Legal Foundation Already in Place
On June 4, 2024, all of Hokkaido was designated as a National Strategic Special Zone for "GX Finance and Asset Management."
Zone policy includes:
- Building GX industry supply chains
- Startup creation/nurturing and advanced technology demonstration/implementation
- Funding environment development
- Attracting domestic and international companies/talent
- Economic revitalization leveraging regional characteristics
"Cold-climate EV operation demonstration and implementation" perfectly aligns with policy items 2 and 5.
5.2 Use the "Legacy" of the 2018 Blackout
September 6, 2018, 3:07 AM. Hokkaido Eastern Iburi Earthquake (M6.7, max intensity 7). Japan's first-ever blackout. 2.95 million households without power. 45 hours to 99% restoration. Complete restoration took about 1 month. Industrial damage: 236.8 billion yen.
One EV (60kWh) can supply 2–3 days of household electricity via V2H. For winter blackout heating security, V2H value is orders of magnitude greater than in Honshu.
Movement already exists. Hiyama Subprefectural Bureau introduced 2 EVs as official vehicles with car-sharing. Rankoshi, Niseko, and Kutchan signed disaster prevention agreements with Nissan for EV power supply during disasters.
Elevating this to Hokkaido-wide institutional design is the Five Arrows.
6. Hokkaido EV Operation Optimization Simulator
A tool for policy makers, automakers, and energy companies to quantitatively analyze winter EV operations in Hokkaido.
import numpy as np
from dataclasses import dataclass
@dataclass
class EVSpec:
"""EV Vehicle Specifications"""
name: str
wltp_range_km: float
battery_kwh: float
has_heat_pump: bool
preconditioning_power_kw: float = 5.0
preconditioning_time_min: float = 20.0
def estimate_winter_range(ev: EVSpec, temp_celsius: float, trip_hours: float = 1.5) -> dict:
"""
Winter range estimation
Physical basis: Arrhenius ionic conductivity, preheating consumption, cabin heating
"""
cond_ratio = conductivity_ratio(temp_celsius)
preconditioning_kwh = (ev.preconditioning_power_kw * ev.preconditioning_time_min / 60)
base_heating_kw = 3.0
effective_heating_kw = base_heating_kw / 2.0 if ev.has_heat_pump else base_heating_kw
heating_kwh = effective_heating_kw * trip_hours
available_kwh = ev.battery_kwh - preconditioning_kwh - heating_kwh
conductivity_loss_factor = 1 - (1 - cond_ratio) * 0.25
effective_kwh = max(available_kwh * conductivity_loss_factor, 0)
energy_per_km = ev.battery_kwh / ev.wltp_range_km
estimated_range = effective_kwh / energy_per_km
loss_pct = (1 - estimated_range / ev.wltp_range_km) * 100
return {"estimated_range_km": round(estimated_range), "loss_pct": round(loss_pct, 1)}
def estimate_charging_time(battery_kwh, temp_celsius, charger_kw=50, soc_start=0.2, soc_end=0.8):
"""Winter fast charging time estimation with BMS throttling"""
charge_kwh = battery_kwh * (soc_end - soc_start)
if temp_celsius < -15: bms_factor = 0.25
elif temp_celsius < -10: bms_factor = 0.35
elif temp_celsius < -5: bms_factor = 0.55
elif temp_celsius < 0: bms_factor = 0.75
else: bms_factor = 1.0
effective_kw = charger_kw * bms_factor
return {"charge_time_min": round((charge_kwh / effective_kw) * 60), "time_ratio": round(1/bms_factor, 1)}
# Simulation: Asahikawa (-7.5°C)
ev_models = [
EVSpec("Tesla Model Y LR", 533, 75, has_heat_pump=True),
EVSpec("Nissan Leaf e+ (62kWh)", 458, 62, has_heat_pump=False),
EVSpec("BYD ATTO 3", 470, 60, has_heat_pump=True),
EVSpec("Leaf (40kWh)", 322, 40, has_heat_pump=False),
]
for ev in ev_models:
r = estimate_winter_range(ev, -7.5)
hp = "HP" if ev.has_heat_pump else " "
print(f"{ev.name:25s} [{hp}] {r['estimated_range_km']:>4}km (-{r['loss_pct']}%)")
(Note: Conceptual simulation for trend analysis and policy discussion, not measured values.)
Vol.1 Summary — The Direction Physics Points Policy
Physical facts established in this paper:
Fact 1: Lithium-ion battery ionic conductivity drops to approximately 49% of room temperature at -20°C (Arrhenius equation calculation).
Fact 2: NAF winter tests measured average -20% range loss at -10°C and -29% to -46% at -31°C.
Fact 3: However, heat pump-equipped vehicles kept losses to under -5% even at -10°C. The range solvable by technology is large.
Fact 4: Low-temperature fast charging carries lithium plating risk, and BMS limits charging speed. Winter charging time is 2–3x that of Honshu.
Fact 5: Hokkaido's renewable electricity has over 1 GW surplus being disposed of through output curtailment.
Fact 6: Hokkaido was designated as a GX National Strategic Special Zone in June 2024, with a regulatory reform proposal foundation already in place.
Policy implications from these facts are clear.
Preview: Vol.2 "Sodium-Ion Batteries — The Game Changer for Cold-Climate Batteries"
Lithium-ion batteries are weak against cold — Vol.1 demonstrated the physical basis. A technology that could fundamentally solve this problem from materials science is now emerging from CATL.
CATL's 2nd-generation sodium-ion battery "Naxtra":
- Maintains 90% discharge capacity at -40°C
- Energy density 200 Wh/kg (+30% vs. 1st gen)
- Fast charge: 80% in 10 minutes (SOC 10→90%)
90% at -40°C. A completely different dimension from lithium-ion batteries that lost at best -29% at -31°C in NAF tests.
Vol.2 will explain sodium-ion battery electrochemistry from first principles and redesign Hokkaido's winter operations from the perspective of materials science.
Series Structure
| Vol. | Theme | Keywords |
|---|---|---|
| Vol.1 (this paper) | Cold-climate battery physics + Special Zone overview | Arrhenius equation, NAF, Five Arrows |
| Vol.2 | Sodium-ion batteries | CATL Naxtra, -40°C 90% retention |
| Vol.3 | All-solid-state batteries | Toyota, sulfide-based, cold resistance |
| Vol.4 | Cold-climate EV operations engineering | Preheating strategy, heat pump, V2H design |
| Vol.5 | Charging infrastructure design | Norway comparison, dead zone elimination |
| Vol.6 | Policy proposal | Five Arrows institutional design, cost estimation |
All articles in this series are published under MIT License.
Citation, reproduction, modification, and commercial use are all free. The purpose of this series is to serve as data and evidence for everyone involved in Hokkaido's EV policy.
This article is a co-production of dosanko_tousan (@dosanko_tousan) and Claude (Anthropic claude-sonnet-4-6).
MIT License — All concepts, code, and frameworks may be freely used, modified, and distributed.
"You can't use EVs in cold places" — We break that assumption with physics and policy. That's this series' job.