THE HYDROGEN INFRASTRUCTURE TRANSITION (HIT) MODEL ----CASE STUDY FOR URBAN BEIJING

摘要

We introduce the Hydrogen Infrastructure Transition (HIT) model and apply it toBeijing, China. The HIT model is a dynamic programming model that generates the spatialand temporal infrastructure buildup decisions that minimize the net present value of capitaland operating costs, carbon taxes, and refueling travel time disbenefits over time. The HITmodel incorporates regionally specific spatial data about road networks, traffic flows andhydrogen demand distribution to find optimal strategies for meeting an exogenouslyspecified market penetration over time. Input assumptions can be varied to test thesensitivity of strategies to technological evolution, feedstock prices, carbon taxes, andmarket penetration rates.We consider 4 scenarios: base case, increasing natural gas prices, rapid technologyimprovement, and rapid market penetration. For each scenario, we show 1) the least-costspatial and temporal decisions generated by the HIT model; 2) the optimal infrastructurelayout; 3) levelized costs over time; 4) well-to-wheel carbon emissions over time.Our findings are as follows: 1) the starting infrastructure configuration for all the 4scenarios during 2010 to 2014 (beginning of the planning horizon) is 30 onsite steammethane reformer (SMR) stations with 3 ton per day per station. These stations serve onlyhydrogen taxis and buses (assuming the government will first introduce hydrogen taxis andbuses) and provide a basis to attract the private fuel cell vehicle purchase, which is assumedto start from year 2015. 2) Regional spatial features have a significant impact on cost. Usinga spanning tree optimization algorithm, we find that the high vehicle density and ring roadnetwork in urban Beijing can be served by a compact pipeline network with a total length ofseveral hundred kilometers. This is shorter than previously reported pipeline designs. 3)Faster market penetration could make a better business case because scale economies inproduction and delivery can be taken advantage of earlier. 4) Carbon policy would need tokeep pace with market penetration to avoid high CO2 emissions from coal gasificationplants without carbon capture technology. If demand increases rapidly, a higher carbon taxmight be needed to drive the adoption of carbon capture technology. 5) Faster technologyimprovement lowers cost. 6) For each scenario, we examine the levelized cost over time fora 12% rate of return. For the base case, the pricing policy of $2.8/kg from 2010 through 2019,$1.8/kg from 2020 through 2059 and $1.1/kg from 2060 onward could achieve a 12% rate ofreturn, ignoring the effect of price on demand.