This page details the regional cost estimate for the Global Gas Infrastructure Tracker (GGIT) that form the basis of capital expenditure projections in our summary tables and briefings/reports. Note that the GGIT is an ever-evolving database, and these cost estimates are updated on a yearly basis with each official release. Because of this, global, regional, and country-level cost estimates vary through time, and efforts are made to make these estimates robust while still representative of the current database at any stage, as described below.

LNG terminals

September 2025 data release

These are the current cost estimates for LNG terminals, superseding the October 2023 values below. The method is unchanged. GEM has identified reported cost estimates for a subset of projects in the GGIT database, collected from media, company websites, and other sources; these are used to estimate the costs of projects for which GEM does not have cost data. Reported costs are normalized by the capacity of each terminal to estimate the cost (in USD) to build a terminal per unit mtpa. The upper and lower 10% of outliers are removed from the distribution to provide a more robust average, and only one cost is used per unique LNG terminal in cases where multiple trains or phases exist. Costs are then grouped by project type — import vs. export and onshore vs. floating — as these different facility types carry different capital expenditure costs. Regional averages are used where a region has at least three data points, and global averages otherwise — applied separately to onshore and to floating terminals. After the trim, the cost sample comprises 178 onshore and 75 floating terminals. These values were restated in September 2026: through the October 2023 release, floating terminals were priced at the global average unconditionally, and a coding error meant the floating rates were not in fact applied to floating units at all — every unit was priced at its region's onshore rate. Both are corrected here. See the table below for the regional cost averages used for extrapolation.


	Import terminals				Export terminals
	Onshore		Floating		Onshore		Floating
Region	Cost (US$ million per mtpa)	Number of data points	Cost (US$ million per mtpa)	Number of data points	Cost (US$ million per mtpa)	Number of data points	Cost (US$ million per mtpa)	Number of data points
Africa	290.3	1	168.9	3	412.6	9	1174.4	6
Americas	384.0	10	162.2	2	703.3	31	755.0	11
Asia	285.8	91	144.6	20	726.9	8	876.9	5
Europe	278.1	17	163.3	19	734.0	6	896.2	1
Oceania	290.3	1	227.3	4	492.1	3	1105.6	3
Global	290.3	120	162.2	48	652.8	57	896.2	26

Cells with fewer than three data points carry the global average rather than their own: onshore imports in Africa and Oceania, floating imports in the Americas, and floating exports in Europe.

October 2023 data release and briefing

Superseded by the September 2025 values above; retained for reference.

GEM has identified reported cost estimates for just over 25% of projects in the GGIT database collected from media, company websites, and other sources. These reported costs are used to estimate the costs of projects for which GEM does not have cost data. To do so, these reported costs are normalized by the capacity of each terminal, to estimate the cost (in USD) to build a terminal per unit mtpa. The upper and lower 10% of outliers are removed from the distribution to provide a more robust average, expansion projects are removed to ensure the costs are based on new builds, and only one cost is used per unique LNG terminal, in cases where multiple trains or phases exist. From there, costs are grouped by project type — import vs. export and onshore vs. floating — as these different facility types carry different capital expenditure costs. For all onshore terminals, regional averages are used if there are at least three data points, and global averages are used otherwise. For all floating terminals, global averages are used. These average cost estimates are used to extrapolate capital expenditure for terminals that do not have reported cost estimates in GGIT. See the table below for regional cost averages used for extrapolation.


	Import terminals				Export terminals
	Onshore		Floating		Onshore		Floating
Region	Cost (US$ million per mtpa)	Number of data points	Cost (US$ million per mtpa)	Number of data points	Cost (US$ million per mtpa)	Number of data points	Cost (US$ million per mtpa)	Number of data points
Africa	386.3	2	165.3	2	679.2	6	669.4	5
Americas	301.8	4	165.3	0	748.6	34	669.4	7
Asia	365.0	22	165.3	13	634.6	7	669.4	4
Europe	444.7	7	165.3	19	1150.1	8	669.4	2
Oceania	386.3	0	165.3	4	770.0	6	669.4	0
Global	386.3	35	165.3	38	783.5	61	669.4	18

Gas pipelines

November 2025 data release

These are the current cost estimates for gas pipelines, superseding the December 2023 values below. The GGIT database has reported cost estimates for just under 30% of gas pipeline projects, which are used to estimate cost projections at the subregional level. The method is unchanged from December 2023: costs are normalized by the length of each pipeline to estimate the cost (in USD) to build a pipeline per unit km, the upper and lower 2.5% of outliers are removed from the distribution to provide a more robust average, expansion projects are removed to ensure the costs are based on new pipeline builds, and the remaining costs are grouped by subregion and averaged. A published subregional average requires at least three pipelines behind it; where a subregion has fewer, the broader average regional cost is used. As of this release that is the case for Melanesia (one pipeline), Micronesia and Polynesia (none), all three of which use Oceania's regional cost per km. Oceania's own regional average is based on Australia and New Zealand's cost data together with that one Melanesian pipeline. See the table below for these costs.

Region	Subregion	Cost (million US$ per km)	Number of data points going into average
Africa		2.67	48
	Northern Africa	2.19	42
	Sub-Saharan Africa	3.78	8
Americas		4.57	267
	Latin America and the Caribbean	2.78	66
	Northern America	5.21	207
Asia		1.80	663
	Central Asia	2.89	28
	Eastern Asia	1.43	551
	South-eastern Asia	2.39	8
	Southern Asia	2.10	23
	Western Asia	3.81	62
Europe		3.20	186
	Eastern Europe	3.00	89
	Northern Europe	3.68	18
	Southern Europe	3.03	61
	Western Europe	3.87	34
Oceania		2.68	19
	Australia and New Zealand	2.80	19
	Melanesia	2.68	1
	Micronesia	2.68	0
	Polynesia	2.68	0

December 2023 data release and briefing

Superseded by the November 2025 values above; retained for reference.

The GGIT database has cost estimates for just over 25% of projects, which are used to estimate cost projections at the subregional level. To do so, these costs are normalized by the length of each pipeline to estimate the cost (in USD) to build a pipeline per unit km. The upper and lower 2.5% of outliers are first removed from the distribution to provide a more robust average, and any expansion projects are also removed to ensure the costs are based on new pipeline builds. From there, costs are grouped by subregion, and an average is taken. In cases where pipeline cost data are not available for a given subregion, the broader average regional cost is used. This is the case for Melanesia, Micronesia, and Polynesia. Oceania’s regional cost per km is used for these subregions, and this cost estimate is based on Australia and New Zealand’s cost data. See the table below for these costs.

Region	Subregion	Cost (million US$ per km)	Number of data points going into average
Africa		3.10	29
	Northern Africa	2.65	26
	Sub-Saharan Africa	3.93	5
Americas		4.31	217
	Latin America and the Caribbean	2.96	43
	Northern America	4.65	179
Asia		1.78	456
	Central Asia	2.28	21
	Eastern Asia	1.43	374
	South-eastern Asia	2.44	8
	Southern Asia	2.26	21
	Western Asia	3.60	41
Europe		3.00	134
	Eastern Europe	2.85	74
	Northern Europe	5.76	10
	Southern Europe	2.36	55
	Western Europe	4.76	13
Oceania		2.85	17
	Australia and New Zealand	2.85	17
	Melanesia	2.85	0
	Micronesia	2.85	0
	Polynesia	2.85	0

Oil pipelines

June 2026 data release

The GOIT database has reported cost estimates for a subset of crude oil and NGL pipeline projects, which are used to estimate cost projections at the subregional level. The method matches the one used for gas pipelines. Costs are normalized by the length of each pipeline to estimate the cost (in USD) to build a pipeline per unit km, the upper and lower 2.5% of outliers are removed from the distribution to provide a more robust average, and expansion projects are removed to ensure the costs are based on new pipeline builds. From there, costs are grouped by subregion and an average is taken. Where a subregion has fewer than five data points, the broader regional average is used; where a region has fewer than five data points, the global average is used. The table below covers crude oil pipelines, and is based on 188 unique pipelines with both a known length and a reported cost. Note that the data points behind several subregions are sparse, and that the figures below therefore carry more uncertainty than the equivalent gas pipeline estimates.

Region	Subregion	Cost (million US$ per km)	Number of data points going into average
Africa		3.19	18
	Northern Africa	2.11	11
	Sub-Saharan Africa	4.11	7
Americas		2.99	50
	Latin America and the Caribbean	2.81	14
	Northern America	3.07	36
Asia		3.07	100
	Central Asia	1.30	6
	Eastern Asia	3.03	68
	South-eastern Asia	3.07	4
	Southern Asia	5.50	8
	Western Asia	3.21	19
Europe		2.54	14
	Eastern Europe	3.50	13
	Northern Europe	2.54	0
	Southern Europe	2.54	4
	Western Europe	2.54	1
Oceania		1.22	6
	Australia and New Zealand	1.22	6
	Melanesia	1.22	0
	Micronesia	1.22	0
	Polynesia	1.22	0

If you need cost estimates not covered here, please contact Global Energy Monitor.

GGIT and GOIT cost estimates

Contents