In the search for clean energy solutions to displace greenhouse gas emitting fossil fuels, few technological options are as alluring as directly producing hydrogen from sunlight.

If hydrogen, the most abundant element in the universe, could be produced on earth economically and with a minimum overall environmental impact, it could provide energy to both stationary and transportation applications with very low overall carbon footprint and climate impact. For example, hydrogen could be the fuel input in fuel cells to generate electricity, or feedstock for producing liquid transportation fuels.

Today however, the most economical way to make hydrogen is by reforming fossil fuels such as natural gas-with the nearly same negative impact to the climate as direct combustion. Hydrogen production via electrolysis-splitting water into hydrogen and oxygen using electricity-can in principle use renewable electricity, but it is currently much more expensive.

Scientists are pursuing a promising pathway to generating large-scale amounts of hydrogen for clean energy production directly by splitting water using sunlight, a process called photoelectrochemical (PEC) production. Instead of splitting off the hydrogen from hydrocarbons and being left with carbon, which is typically oxidized and emitted into the atmosphere as carbon dioxide, photoelectrochemical production splits off hydrogen from water, leaving clean oxygen gas.

Researchers have accomplished PEC on a small scale in laboratories, but scaling up the process into hydrogen generating plants capable of supplying enough to meet the needs of industrial societies requires considerably more research and technology development.

Many unanswered questions lie not just in the technology, but in the area of life-cycle impact-in particular, its net energy balance. An energy production facility such as one based on PEC technology, should generate more energy over its lifetime than is used to manufacture and operate it. Scientists and funding agencies would like to understand what research directions they need to follow in order to make large-scale PEC-based hydrogen production a reality.

A new study from scientists at the Joint Center for Artificial Photosynthesis (JCAP) created a life-cycle assessment (LCA) model to provide some estimates that might help guide research directions to faster marketplace success. They constructed a model simulation of a large-scale PEC-based hydrogen production facility, using what is known currently about the technology as well as projections of future performance.

JCAP scientists affiliated with the Materials, Physical Biosciences, Chemical Sciences and Environmental Energy Technologies (EETD) Divisions of Lawrence Berkeley National Laboratory (Berkeley Lab) participated in the study.

"The modeling of this solar-to-hydrogen technology provides insights into its potential competitiveness," says the study's lead author, Roger Sathre of the Environmental Energy Technologies Division. "It will help identify the key challenges and opportunities for improvement."

EETD researchers have had considerable experience performing life cycle assessments of technologies still in the laboratory, such as new infrared-blocking electrochromic window coatings, carbon sequestration technologies, as well as advanced biofuels. Their results are intended to help guide lab R and D to market success.

Thorough description and many inputs
The development of the hydrogen production model required many components, and considerable input from the researchers developing the technology. The research team modeled a facility capable of producing the hydrogen equivalent of 1 GW of continuous output, or 610 tons of hydrogen per day. All U.S. light-duty vehicles could be powered by about 160 such plants.

"This study is the first to look at a large hydrogen generation system, and to make a thorough assessment of its balance of system [BOS] requirements-its energy and materials inputs and outputs," says Jeffery Greenblatt of EETD, one of the study's authors. A couple of prior studies have evaluated smaller scale systems, about one-thousandth the size, focusing on their economics.

The Berkeley researchers prepared a preliminary engineering design of the plant, and generated a model describing the system-wide energy flows associated with producing, using, and decommissioning the facility. This allowed them to calculate the facility's three primary energy metrics.

One is the life-cycle primary energy balance, or how much net energy the facility would provide over its lifetime. The second is the energy return on energy investment (EROEI), which describes how much usable energy the facility generates divided by its energy inputs-it must be greater than one by as much as possible for the technology to be viable.

Finally, the energy payback time measures how long the facility must operate to deliver the hydrogen equivalent of the energy required for its manufacturing, construction and decommissioning.

Creating the model required building the facility up from its components. The team modeled a large, two square meter photoelectrochemical cell, assembled in a truck-transportable structure called a panel containing 14 cells. Panels were arranged in fields consisting of 1,000 panels each, and the overall facility was made up of 1,510 fields.

The model required estimates of energy use to make all these components, plus the rest of the plant such as pipes for water and gas, storage tanks, compressors, sensors, roads, and the like. Construction, operation and decommissioning required estimates of the energy inputs, material inputs such as water and process gases, and transportation to bring in materials and cart out wastes. The plant was assumed to have a service life of 40 years.

Positive energy benefits require meeting system-level goals
Under the model's base case conditions, the plant's payback time is 8.1 years. The energy return on energy invested, at 1.7, is positive. The life-cycle primary energy balance over the plant's 40-year life is more than 500 petajoules. "One petajoule is the energy required to power 50,000 hydrogen fuel-cell cars for a year," Greenblatt points out.

"Our results show that hydrogen production based on photoelectrochemical technology has the potential to deliver significant amounts of energy," says Sathre. "There are a number of variables that influence how much energy, and these are variables that R and D in the field needs to focus on."

The most important factor is the overall efficiency of conversion from solar energy to hydrogen, termed the solar-to-hydrogen (STH) efficiency ratio. The higher the STH efficiency, the better the energy return (the base case assumed 10 percent conversion efficiency).

The lifespan of the PEC cell, the energy used to manufacture the PEC cell, and the lifespan of the rest of the facility are the other most important factors. The report addresses a number of research directions that could lead to more efficient PEC cells.

The researchers estimate that if PEC cells have an STH efficiency of 20 percent (which they believe is possible eventually), and a cell life span of 20 years, the plant can have an energy payback time of just three years, and an EROEI of more than 3, almost double that of the base case.

"Our result validates the need for high efficiency PEC cells, something the research community already understands," says Frances Houle, Department Head for Science-Based Scale-Up at JCAP and another of the study's authors.

"It also drives home the need for cell longevity-on the scale of years-well beyond what is currently measured in the lab, which is the scale of hours. Also, we found that the energy investment in the balance of system is smaller than that required to fabricate the PEC cells, so methods to make the cells with less energy will be impactful."

Greenblatt adds that "research is on the right track, because the analysis suggests that a plant built with PEC technology will be energy-positive, but future R and D should ensure that the variables most affecting net energy balance-efficiency, longevity, initial energy investment-are well-understood and optimized."

Roger Sathre, Corinne D. Scown, William R. Morrow III, John C. Stevens, Ian D. Sharp, Joel W. Ager III, Karl Walczak, Frances A. Houle, and Jeffery B. Greenblatt. "Life-cycle net energy assessment of large-scale hydrogen production via photochemical water splitting," Energy and Environmental Sciences, DOI: 10.1039/c4ee01019a.