Commentary: Battery storage is (almost) ready to play the flexibility game



7 February 2019

Lithium ion batteries - mostly aimed at providing short-term storage - now account for just over 80% of all battery capacity (Photograph: Shutterstock)

In an electricity world that sees variable renewables at the centre stage, market players and policy makers cannot overlook the need for flexibility, a game where storage technologies are expected to play a key role.

The strong expansion of variable renewables, which are projected to make up more than half of global capacity additions to 2040 in the IEA’s New Policies Scenario (NPS), has major implications for electricity, first among them is the need for increased flexibility. Globally, electricity demand is projected to grow by over 20% over the next decade, but flexibility – the ability of the power system to quickly adapt to changes in power supply and demand – grows by some 80%. Flexibility will therefore be the cornerstone of future electricity systems. It will be met not only by traditional sources of flexibility – such as conventional power plants and electricity grids – but also by new sources of flexibility, including battery storage and demand-side response, which are projected to grow fast and contribute 400 GW by 2040.

Storage in particular, is attractive to power generators as it leads to higher overall utilisation of power system assets. This translates into a lower risk of overcapacity and higher average revenues.

Today, pumped hydro storage systems account for the majority of storage capacity (153 GW, equivalent to about 2% of total power capacity worldwide), while battery storage systems total around 4 GW. However, while pumped hydro storage is projected to grow in the next decade, the technology deployment is largely constrained by the location of suitable sites.

On the other hand, battery storage systems, which are modular, allow a wide range of applications. As costs continue to fall, installations have tripled in less than three years, largely driven by lithium ion batteries - mostly aimed at providing short-term storage - which now account for just over 80% of all battery capacity. For applications with longer storage durations other battery types, including sodium sulfur and in particular flow batteries, are attracting increased interest. Small-scale battery storage is also making inroads, and in off-grid solar applications for energy access, the vast majority of systems now include a storage unit.

Further cost declines for battery storage systems are expected: costs for four-hour battery systems are projected to fall to $220 per kWh by 2040 in the NPS. Together with appropriate market design that rewards these flexible assets, these falling costs are projected to drive the strong deployment of batteries, with utility-scale deployment reaching 220 GW by 2040 in the NPS. Most battery additions are expected to be paired with solar PV and wind power as they increase their dispatchability, and allow revenue stacking from energy arbitrage and ancillary services offered to the grid. 

Deployment and costs of utility-scale battery storage systems in the New Policies Scenario
	China	India	United States	European Union	Africa	Rest of world
2020	6	5	5	1	1	5
2030	30	27	14	7	6	22
2040	48	61	34	14	15	46
{
	"chart":{
		"type":"column"
	},
	"plotOptions":{
		"column":{
			"borderWidth":0,
			"stacking":"normal"
		}
	},
	"title":{
		"text":""
	},
	"tooltip":{
		"valueSuffix":" GW"
	},
	"subtitle":{
		"text":"Installed capacity"
	},
	"yAxis":{
		"title":{
			"text":"GW"
		},
		"reversedStacks":false
	},
	"xAxis":{
		"type":"category"
	}
}
	Battery pack	Non-battery costs	1 hour	8 hours	4 hours
2017	193	209	619	339	402
2025	128	141	415	226	268
2030	118	124	372	205	243
2035	114	114	347	194	228
2040	111	107	329	186	218
{
	"chart":{
		"type":"areaspline"
	},
	"plotOptions":{
		"areaspline":{
			"marker":{
				"enabled":false
			},
			"stacking":"normal",
			"lineWidth":0
		},
		"spline":{
			"marker":{
				"enabled":false
			},
			"dashStyle":"longDash"
		}
	},
	"title":{
		"text":""
	},
	"tooltip":{
		"valuePrefix":"$",
		"valueSuffix":" per kWh (2017)"
	},
	"subtitle":{
		"text":"Capital cost"
	},
	"yAxis":{
		"title":{
			"text":"dollars per kWh (2017)"
		},
		"reversedStacks":false
	},
	"series":[{
		"colorIndex":2
	}, {
		"colorIndex":3
	}, {
		"type":"spline",
		"colorIndex":7
	}, {
		"type":"spline",
		"colorIndex":8
	}, {
		"type":"spline",
		"colorIndex":2
	}]
}

Another factor set to drive the battery storage boom is the need for peaking capacity, which is projected to increase by three-quarters globally to 2040 in the NPS, and for which stand-alone batteries become competitive on a cost and value basis in many regions in the short term. For example, battery storage becomes competitive with open-cycle gas turbines in India soon after 2020. Meanwhile in the United States, batteries close in on gas turbines near 2030.

What if battery storage becomes really cheap?

In WEO 2018 we assessed the impact of cheap batteries on global power systems, assuming the widespread availability of second-use batteries, and a best-in-class reduction in battery system costs comparable to those experienced in recent years by solar PV systems.

In the NPS cost reductions achieved in batteries for transport would spill over into power sector applications, driving utility-scale battery pack costs to fall to around $100 per kWh by 2030. But a large number of batteries could be re-purposed after use in an electric vehicle for a second life in the power sector: the reduction in energy storage capacity in a battery that would reduce the range of an electric vehicle to the point where a new battery was needed would not prevent the battery from being useful in grid-scale applications. The availability of second-use batteries and further balance of system cost reductions would give a further boost to the competitiveness of battery storage.

	Gas turbines	Oil turbines	Battery storage
	338	235	3
	633	170	218
	532	169	544
{
	"chart":{
		"type":"column",
		"height":"45%"
	},
	"plotOptions":{
		"column":{
			"borderWidth":0,
			"stacking":"normal"
		}
	},
	"title":{
		"text":"Peaking capacity by technology in 2017 and 2040 in the New Policies Scenario and cheap battery case"
	},
	"yAxis":{
		"reversedStacks":false,
		"title":{
			"text":"GW"
		}
	},
	"xAxis":{
		"categories":["2017","2040: New Policies Scenario","2040: Cheap storage case"]
	},
	"tooltip":{
		"valueSuffix":" GW"
	},
	"series":[{
		"colorIndex":2
	}, {
		"colorIndex":9
	}, {
		"colorIndex":8
	}]
}

Under these assumptions, cost reductions would lead to batteries being 70% less expensive than today by 2040, and to battery storage becoming more competitive than alternative options for flexibility earlier than in the NPS. This would translate into 540 GW of batteries installed by 2040, reducing gas turbines by 100 GW and making battery storage the main technology for peaking capacity by 2040.

This would also provide cost savings by avoiding overcapacity in the system and by reducing or deferring the need for some grid infrastructure investment. Finally, batteries paired with variable renewables could further boost renewables deployment through the increased value proposition of these technologies used in combination.

Latest articles