Module 4

Peak-shaving LNG plants

Overview

A gas utility's insurance policy. Liquefy slowly for most of the year, store a big tank of it, then vaporise it fast on the handful of days when everyone's heater is on and the pipeline can't keep up.

Peak-shaving — nameplate
OwnerGas distribution utilities and pipeline operators, not producers
LiquefierSmall — 50–500 tpd — runs 150–250 days a year to fill the tank
StorageOne or two large tanks, 20,000–100,000+ m³. The tank is the point of the plant
Send-outVaporisers sized 10–30 times the liquefaction rate — thousands of tonnes per day for a few days
RefrigerationNitrogen expander or SMR; older plants used cascade or mixed refrigerant
Duty cycleFill for months, hold, send out for 5–20 days a year — sometimes zero
StandardNFPA 59A (US), EN 1473 (Europe), AS 3961 for LNG storage and handling in Australia
Loading the peak-shaving process deck…

Why peak-shaving exists

Gas demand in a city with winter heating isn't flat. Melbourne's daily gas use on a cold July morning can be several times its January use. The pipeline into the city is sized for something between the two, because building a pipeline to carry the peak day would leave it mostly empty. Contracting extra pipeline capacity for those few days is expensive — capacity is charged whether you use it or not — and buying gas on the spot market on the peak day is when it costs the most.

A peak-shaving plant solves this by time-shifting. Gas is bought cheap and liquefied slowly in summer when the pipeline has spare capacity, stored as LNG, and vaporised back into the network on the peak days. The "shaved" peak is the part of the demand curve above the pipeline capacity. In the US there are about a hundred of these plants, mostly in the northeast and midwest; the UK, Germany and Australia have a handful.

The three-part plant

Unlike every other plant type, a peak-shaver is designed around the tank and the vaporisers. The liquefier is the smallest, least important-looking bit.

  1. Liquefier (fill mode, most of the year). Gas is taken off the distribution network, treated (CO₂, water, mercury, heavies) and liquefied at a modest rate — 100 tpd would take about 200 days to fill a 50,000 m³ tank. Nitrogen expander or SMR cycles are standard because they're simple to start and stop.
  2. Storage (hold mode). A single-containment or full-containment tank holds the inventory for months. Boil-off from heat leak (typically 0.05–0.1% of contents per day) is compressed and sent back into the gas network, so nothing is lost — the network is a permanent sink for BOG. Managing stratification and rollover over a long hold is the main storage concern.
  3. Vaporisation (send-out mode, a few days a year). In-tank pumps lift LNG to network pressure; high-pressure vaporisers heat it back to gas. On a peak day the send-out can be 5,000 tpd or more from a plant whose liquefier makes 100 tpd. The gas is odorised, metered and pushed into the distribution network at the pressure the network needs.

Vaporiser types

TypeHow it worksSuits
Submerged combustion vaporiser (SCV)A gas burner fires into a water bath; LNG flows through a coil in the bath. Fast start, very high capacity, burns about 1.5% of the gas sent out.Peak send-out — the workhorse
Shell-and-tube with water/glycolWarm water-glycol from a boiler or waste heat source heats LNG in a conventional exchanger.Steady send-out, where waste heat exists
Ambient air vaporiser (AAV)Tall finned tubes exposed to air. No fuel, but ices up and needs switching between banks; capacity drops in cold weather — exactly when you need it.Small plants and warmer climates; often used as the first stage before a trim heater
Open rack vaporiser (ORV)Seawater runs down panels of finned tubes.Coastal import terminals more than peak-shavers

Satellite peak-shavers

Some plants skip the liquefier entirely. They're just a tank and vaporisers, filled by road tanker from a small-scale or baseload plant elsewhere. This is the cheapest way to add peak capacity to a regional network. In Australia, the same tank that serves a small-scale trucking business can double as a satellite peak-shaver for a regional town's gas supply.

Operating a peak-shaver

The operating year has a rhythm that no other plant type has:

  • Autumn: tank should be full. Vaporisers are tested, burners proven, pumps run up, the send-out header pressure-tested. Readiness is everything — the plant will get one or two calls and must respond within an hour.
  • Winter: mostly waiting. On a forecast peak day the network control room calls for send-out at a rate and time. The plant ramps vaporisers, holds the rate for hours or days, and ramps down. Between calls it's back to monitoring boil-off.
  • Spring and summer: the liquefier runs to refill what was sent out, plus the boil-off lost over winter. Maintenance is done on the vaporisers while they're idle.

Because the liquefier only needs to refill the tank once a year, it runs for long stretches at a steady rate and then sits idle for months. That start-stop life is why nitrogen cycles and simple SMR loops are preferred — no propane inventory to manage, no refrigerant composition to nurse.

The Australian picture

Dandenong LNG in Melbourne's south-east is the main example. It was built in the 1980s to peak-shave Victoria's winter demand and has since grown a road-tanker business supplying LNG to industrial customers and transport. Its role in the Victorian gas system is written into the market rules: the operator can call on it when transmission system pressure is under stress. Other Australian LNG storage tied to peak demand tends to be at import terminal proposals on the east coast, which are a different animal — they import LNG by ship and vaporise it, with no liquefier at all.

Economics in one paragraph

A peak-shaver earns its keep by avoiding two costs: firm pipeline capacity reserved for a peak that only shows up a few days a year, and spot gas bought at the top of the market. Against that sits the capital cost of a large cryogenic tank and the fixed cost of keeping a plant on standby that might not run at all in a mild winter. Utilities justify it as reliability — the alternative to a peak-shaver is curtailing industrial customers or, in the worst case, losing pressure in the residential network, which is a multi-week relight nightmare.

Safety specifics for peak-shaving

Peak-shavers sit close to cities — that's the point — so the tank design, thermal radiation and vapour dispersion exclusion zones are the dominant design constraints. Long storage holds raise the rollover risk: LNG of slightly different density layers in the tank, the bottom layer warms, and when the layers suddenly mix a large burst of vapour can overwhelm the relief system. Modern plants monitor tank temperature and density profiles and recirculate to prevent stratification. Vaporiser start-up on a cold morning after months idle is the other high-risk moment: burner light-off, cold LNG hitting warm metal, and pressure control on a network that's already under stress.

Self-check — Module 4
  1. What does "shaving the peak" mean in terms of the demand curve and the pipeline?
  2. Why is the liquefier tiny compared to the vaporisers?
  3. What is a satellite peak-shaver?
  4. Explain rollover and why a long storage hold makes it more likely.
  5. Which vaporiser type burns some of the product to work, and roughly how much?