NEDO Microgrid Case Study

Case study: －U.S. Collaborative Smart Grid

Demonstration Project in New Mexico

Part 2 Efforts in Albuquerque

Hiroshi Irie（Mitsubishi Research Institute）

1. Introduction contribute to the operation of electric system has been conducted. For the period between FY2009 and FY2014, the Japan – U.S. The BEMS constructed by incorporate two Collaborative Smart Grid Demonstration Project in New Mexico was control algorithms by Shimizu Corporation and Gas, which are implemented at two sites, Los Alamos and Albuquerque in New called here BEMS1 and BEMS2, respectively. Both BEMS algorithms Mexico State in the United States as NEDO’s first Overseas Smart controlled the same equipment and devices but had different purposes. Community Demonstration Project. Headed by NEDO, the project was In this case study, the major efforts made in the site of Albuquerque are accomplished by a collaboration of 19 Japanese companies, the State described from the two BEMSs perspectives as follows: Government of New Mexico, electric utilities, research organizations, and other stakeholders demonstrate technologies that allow for large  BEMS1: Contribution as microgrid as a whole to the operation of scale penetration of renewable energy in the future, and offering some electric system important clues on the promotion of smart grid in both Japan and the  BEMS2: Coordinated operation with other control system U.S. in the course of this demonstrative experiment. This Document is a case study focusing on the efforts deployed in Albuquerque and report on the suggestions made by this experience.

2. Efforts in Albuquerque In Albuquerque, an advanced microgrid consisting of Building Energy Management Systems (BEMSs) were constructed in a commercial building called the Aperture Center located at the Mesa Del Sol residential development. A research about how to operate the Source: Photo taken by author microgrid to optimize the facility’s energy supply and possibly Figure 2 Aperture Center building at Mesa Del Sol

Efforts in Los Alamos

PV Monitoring

Monitoring & control  State of New Mexico Battery NEDO, , , Shimizu  Los Alamos County Smart house Corp., , Sharp, NGK  Los Alamos National μEMS Insulators, NEC, Meidensha, Fuji Laboratory Load control signal Control Smart Electric, , MHI, Furukawa  Sandia National HEMS appliances, etc Electric, CDI, CTC, , Laboratories PLC Kandenko, NTT Facilities,  Univ. of New Mexico transfer Accenture  PNM Interruption signal trip  Mesa Del Sol General houses (900 homes) Smart Demand response signal/ meter Measure power consumption

Efforts in Albuquerque

Load control signal Commercial building Storage μEMS （MESA DEL SOL） battery Data measurement BEMS

Controller BEMS1 Gas engine

Monitoring Large capacity BEMS2 & Control PV PV

Large capacity battery Data measurement/connection Fuel cell Utility’s PV site

Figure 1 Whole picture of Japan-U.S. Collaborative Smart Grid Demonstration project in New Mexico - 1 -

NEDO Microgrid Case Study

3. BEMS1： Contribution to power system as a one day is shown in Figure 5. For two years, various demonstrative tests microgrid were conducted with real loads, followed by the adjustments of 3.1 System Overview equipment to resolve operational issues that were encountered. As a The first effort is the BEMS1 constructed in the Aperture Center at result, the accuracy of power flow control at the PCC has been Mesa Del Sol by Shimizu Corporation. As shown in Figure 3, the enhanced to approximately 2 kW of standard deviation of PCC power system consists of the following DERs (Distributed Energy Resources): flow over the control target value. a gas-engine generator (240 kW), phosphoric-acid fuel cells (80 kW), a photovoltaic power generator (50 kW), storage batteries (160 kWh) Active power control Target value of Target value of Target value of purchasing power power source 2 power source 3 managed by a Battery Energy Storage System (BESS), and an [kW] [kW] [kW] LPF Limiter 1 Load power[kW] - - - Active Power power air-cooled refrigerator (70 USRT), an absorption type refrigerator (20 LPF1 source + + + 1 3 USRT), a water thermal storage tank (75 m ) and a hot water thermal Active power of Target value of power source 1 power source 3 3 [kW] [kW] LPF Limiter 2 storage tank (110 m ) as heat source systems. Controlled by the BEMS1, - - Active power Power LPF2 source the equipment was configured to supply electric power and thermal + + 2 Active power of Active power of power source 1 power source 2 energy to a commercial building. [kW] [kW] LPF Limiter 3 - - Active Power power The BEMS1 has two major capabilities: (1) control of DERs in the LPF3 source + + 3 system when it is grid-connected so as to stabilize power flow at the Reactive power control Battery Target reactive point of common coupling (PCC) to any externally-set specific value reactive power power at PCC LPF [kVar] [kVar] Limiter 7 Reactive power - - reference value PCC [kVar] Storage (grid-connected mode); and (2) islanded operation of the building at LPF7 Reactive power[kVar] battery + + times of grid outage, enabling switching between interconnected and islanded operation modes as needed without any instantaneous Figure 4 Control method for grid-connected mode interruption (islanding mode). Figure 4 shows the control logic under grid-connected mode. Here, Islanding mode used an algorithm to supply power and heat for an the system was operated with optimal use of each DER, based on its entire commercial building when the circuit breaker at the PCC was performance characteristics by setting specific values in cascade to open so that the building does not receive power supply from the grid. stabilize the power from PCC. The performance of BEMS1 during See Figure 6 for the logic.

Air-cooled Air-cooled Estimated load of building Thermal storage tank refrigerator refrigerator Cooling tower 400kW 70USRT 20USRT

500kW PV 2MWh Batteries

Substation

BEMS Feeder Power controller Heat controller PCC

Smart meter μEMS+MDMS

PV 50kW Dummy load Lead storage battery Fuel cell Gas engine generator 100kW Constructed power line 50/90kW, 160kWh 80kW 240kW PV Inverter Communication line 50kW Existing power line

Figure 3 BEMS configuration

- 2 -

NEDO Microgrid Case Study

PCC BAT PV GE FC LOAD 3.2 Key Findings – Lessons Learned 300 2013.09.09 By the above-mentioned demonstration with BEMS1, it was 250 LOAD confirmed that the microgrid had expected performance as Dr. Kimio 200 Morino of Shimizu Corporation mentioned. Gas Engine 150

100 FuelCell Comment by Dr. Kimio Morino

50 Engaging in the construction of demonstrative system for two years

Power [kW] Power PV 0 from 2010 and the demonstrative experiment for the following two Battery -50 years, the 9 Japanese companies participated in the demonstration PCC -100 project in Albuquerque were able to achieve their original aim with a PCC reference 0kW -50kW -100kW -50kW great support by the U.S. counterpart. -150 10:30 10:45 11:00 11:15 11:30 At our company, we gained huge confidence as we could construct Figure 5 Performance results in grid-connected mode microgrid in the U.S. and successfully demonstrate various microgrid control technologies such as switching between islanding and

LPF limiter 5 Active power control Lower limit [kW] grid-connected modes without any instantaneous interruption using LPF Limiter 4 Load power[kW] - LPF4 BEMS, DR control in coordination with grid, power fluctuation + compensation by use of refrigerator, etc.

LPF Limiter 5 Active power - reference value [kW] Power LPF5 source + 1 The following four lessons have been learned from this Power source 1 active power [kW] LPF experience: Limiter 6 Active power - - reference Power value [kW] LPF6 source + + 2

Power source 1： Fuel cell Power source 2： Storage battery Lesson (1): Microgrid that contributes to the operation of power system GE Generator Reactive power control Target reactive power [kVar] First of all, the BEMS system developed here is able to contribute Battery reactive power LPF [kVar] Limiter 8 Reactive power - - reference GE Generator Storage to the operation of power system on a microgrid-wide basis. When Reactive power LPF8 value [kVar] battery [kVar] + + assuming a building as a microgrid, the power flow at PCC in the Figure 6 Control method for islanding mode microgrid is controlled to become the pre-set target value in the grid-connected mode. In this demonstration, the target value at PCC Figure 7 shows the demonstration result in Islanding mode. In could be changed on a minute-to-minute basis and DERs in the building islanding mode, voltage fluctuation achieved the target of 480V±10% were controlled to achieve the target value in accordance with the target for all operating hours, and frequency stayed in the target range of value. 60Hz±0.3Hz more than 98% of the time, resulting in the stable islanded This means that the microgrid, when seen from the outside, can be operation while maintaining target power quality for long hours. treated as a controllable resource, with the DERs managing the target value and internal load fluctuations. In other words, the system that

PCC BAT PV GE FC LOAD has been developed enabled the microgrid to respond to its load and also 350 Grid-connecting function as a virtual power plant (VPP). Mr. Jonathan Hawkins from the 300 → Islanding Islanding → Grid- connecting LOAD local utility PNM commented on how the demonstration experiment is 250 significant for electric utilities. 200

150 Gas Engine Comment by Mr. Jonathan Hawkins 100 Fuel Cell PCC We were involved in the demonstration because we wanted to clarify Power [kW] Power 50 PV how microgrid and electric utility relate to each other. It is important to 0 deepen understanding of this system which is expected to prevail first -50 Battery and we have to find out how utilities can support microgrid and what -100 kind of coordination would be needed between utilities and microgrid. -150 13:00 13:15 13:30 13:45 14:00 14:15 14:30 14:45 15:00 15:15 15:30 In this respect, we could learn a lot from the demonstrative experiment, Figure 7 Performance results in Islanding mode and we would like to continue to further this research.

- 3 -

NEDO Microgrid Case Study

One point we should keep in mind is that the system has the ability microgrid because it may shift into islanded operation mode. Because to operate as a microgrid and as a VPP as well. However, the system for PNM wanted to clarify this point, they actively participated in the this demonstration has characteristics of contributing to the power demonstration project from the beginning. system operation in the grid-connected mode and also shifting to the It was demonstrated that when under islanded operation mode, the islanding mode without any instantaneous interruption at the time of microgrid has to disconnect itself from the grid at PCC to suspend power grid outage. Today, energy system resilience is an important power supply to the grid and also has to increase short-circuit capacity consideration in the U.S. and Japan. It is significant that the system of the switch at PCC. I remember that we have carried out intensive developed for the demonstration incorporated two features that support studies to implement these functions which were very challenging that energy efficiency as well as resilience. NEDO and all the companies involved faced tremendous difficulties. Mr. Manuel Barrera from Mesa Del Sol, developer of the demonstration site building, described his feelings about the microgrid As explained in the above comment, there was a high bar and the demonstrative experiment. established in the introduction phase, but the project partners were able to meet these requirements and build the system which satisfied both the Comment by Mr. Manuel Barrera utility and the system developer to start the demonstration experiment The demonstrative experiment was really a great experience for us to and obtained fruitful results. IEEE 1547, the standard for grid rediscover the value of microgrid. I believe a property value is improved connection of distributed generation in the U.S., is currently being with the microgrid and I am sure that even with a higher initial revised and the discussions on the standard including the topic of investment the long-term value of the property will be guaranteed. microgrid are underway. The experience at the Aperture Center will be As a developer, we would like to make an effective use of the results of referenced for the future development of standards for the this demonstration to continuously provide the customers who have grid-connection of microgrid. interests in sustainability with our attractive service. Currently, shared use of PV and storage battery at the community level Lesson (3): Implications for the microgrid system design is prohibited by regulation in New Mexico. However, the experiment The BEMS1 system developed for this demonstration is a hybrid this time demonstrated that such a scheme would be possible in an system consisting of a number of energy resources with different actual system. I am confident that the demonstrative experiment can be features. There are various constraints in designing a hybrid system. The a supportive evidence to change the regulation. BEMS control system established in this demonstration project offers a suitable control method to implement optimum operation while The demonstration project was intended to technically verify the overcoming the constraints. effectiveness of a building-scale microgrid. However, as pointed out by The step response characteristics of DERs adopted in the Mr. Barrera, it should be necessary to take an approach toward demonstration project are shown in Table 1. Based on this table, it expansion of microgrid business based on the demonstration results (e.g. may seem critical to focus on the designing of the BESS with the fastest encouraging change in the policy and regulatory frameworks). response. However, the cost of BESS is quite high, so it is desirable to minimize the capacity. Lesson (2): Development of standards for grid connection of a Amid discussions on a trade-off between cost and performance, microgrid the “Cascade Control System” has been implemented in the Microgrid as emerging technology is still in the evolution phase. demonstration project. That is, the BESS can compensate power There exist guidelines such as IEEE 1547.4 but the requirements on the fluctuations that cannot be handled by other generation resources. The connection between the electric system and microgrids are under debate. capacity requirement of costly BESS was minimized by using the No standard on the grid connection of microgrid has been developed BESS for only quick power deviations that could not be corrected by yet. the other DERs. Under such circumstances, we still had to examine the Another important consideration was control of active power when requirements at the PCC. Mr. Jonathan Hawkins of PNM Resources there was not enough power for the microgrid. In the BEMS1 Cascade looked back at that time and commented as follows: Control, active power of each DER was designed to be controlled based on real-time active power measurements. However, the load of the Comment by Mr. Jonathan Hawkins building--specifically, the elevators—became an issue. It was found that The U.S. electric utilities normally request compliance to IEEE 1547 the power flow at PCC fluctuated largely because of the inrush current for PCC, but it is very difficult to define the requirements for PCC for due to the operation of elevators.

- 4 -

NEDO Microgrid Case Study

Table 1 Response characteristics of DERs The quality of service provided by the refrigerator is an important Response Time constraint. The refrigerator was not introduced in the system to assist DER Ramp Rate (Approximately) with the operation of electric power system, but rather to maintain

10min temperature in the building within a target range. If this function of Fuel cell 0.064kW/sec (40kW→80kW) refrigerator is hindered, a trade-off between contribution to the operation of power system and maintaining service quality must occur. 5min 30sec Gas engine 0.45kW/sec (90kW→240kW) 30msec w/o chiller w/ chiller BESS 3000kW/sec 30 (0kW→90kW) 25 4min 30sec Refrigerator 0.15kW/sec (30kW→70kW) 20 0.296kWh

15 To address this issue in the demonstration, a reactive power control

Battery output[kW] Battery 10 function in addition to the active power control function was added to 0.190kWh the BESS, improving the voltage performance. 5 In addition, there were various other lessons learned from the 0 demonstrative experiment. Prof. Andrea Alberto Mammoli of the 0 10 20 30 40 50 60 70 80 Time [sec] University of New Mexico talked about the significance of the project Figure 8 Effect of BESS capacity reduction due to integration of as follows: refrigerator to BEMS

Comment by Prof. Andrea Alberto Mammoli In this regard, there were interesting findings. Figure 9 indicates the A highly complicated system like microgrid is often quite difficult to measurement value of the refrigerator’s temperature regulation when explain only by the theory but can rather be understood by actually integrated into BEMS1. The results indicate that neither TCOP building and operating the real system. (operational efficiency) nor cold water outlet temperature were For the aspects of microgrid design, maintenance and operation, we significantly affected. In general, thermal energy has a long time came up with quite many facts that we could only understand by constant. That is why a refrigerator can be integrated to a power system actually doing it. In that sense, the demonstrative experiment has as a controlled device without losing service quality. It is highly important implications and we have to make a further leap forward desirable to make the best possible use of operational flexibility based on the knowledge obtained from the project this time. provided by thermal management devices.

TCOP Flow rate (1min Ave) Inlet temp Outet temp 6 35 Lesson (4): Capability of thermal load for the stabilization of power Chiller + Battery , Oct 23 2013 5 30 system TCOP

In this demonstration project, a refrigerator was used as thermal 4 25 Operated with

load to demonstrate the capability of thermal load for the stabilization of high COP ] ℃ 3 20

power system. By externally controlling the power consumption of this Inlet temp

], Flow rate [m3/h] rate ], Flow

Temp. [ Temp. - refrigerator, we were able to integrate the refrigerator into the microgrid 2 15 Outlet temp as a demand response device. [ TCOP 1 10 The overall control performance is enhanced with the integration No drastic disturbance in cool Flow rate of refrigerator as a demand response device. In this manner, the power 0 water outlet temperature 5 09:30 09:40 09:50 10:00 10:10 10:20 10:30 consumption can be controlled up to a certain level; therefore, the Figure 9 TCOP and cold water outlet temperature of the refrigerator required capacity and cost of the BESS could be minimized. Figure 8 which is integrated and controlled shows the data of BESS’ reduced capacity resulted from the integration of the refrigerator. In this case, a 30% reduction in the BESS capacity was achieved.

- 5 -

NEDO Microgrid Case Study

4. BEMS2: Coordinated operation with other control The large energy storage system (battery) was installed by PNM system on the grid side to compensate the PV plant fluctuations. However, as 4.1 System Overview previously mentioned, storage batteries are expensive. The second effort is the BEMS project led by Tokyo Gas, which The demonstration was intended to demonstrate that the capacity was also constructed in the Aperture Center at Mesa Del Sol. BEMS1 of of the large batteries could be reduced by coordinating the PV Shimizu Corporation and BEMS2 of Tokyo Gas are different systems smoothing control with a gas engine and fuel cell using BEMS2. This implemented in the same hardware. experimental demonstration was implemented by Tokyo Gas and The configuration of the entire BEMS2 control system is shown in Sandia National Laboratories with the cooperation of Shimizu Figure 10 and Figure 11, and consisted of the same DERs as Shimizu’s Corporation, Toshiba and PNM. BEMS1 with a gas engine, fuel cell, PV, batteries and thermal source The control logic of this BEMS2 is shown in Figure 12. In this system. There were also two operation modes of (1) grid-connected scheme, the output fluctuations of the PV plant is compensated by two mode and (2) islanding mode in this Tokyo Gas’ BEMS2, the same two generation sources of gas engine and fuel cell, and the difference in the modes implemented in BEMS1. However, Tokyo Gas took the output of PV plant and gas engine is compensated by the large storage approach to place particular importance on supporting the operation of a batteries installed on the grid side. large-scale PV plant (500 kW) and the large-scale energy storage (lead batteries, 500 kW/500 kWh) owned and operated by the electric power Lower PV Power High Variability PV Power (PPV) Variability company (PNM). The large-scale PV plant and battery is located Σ Production Power P PV Battery P Battery P (PPV + PGE + Pbat) approximately 2 km from the Mesa Del Sol site, and are electrically Sensor Bat-SP Bat Controller Plant connected to the same feeder as the Aperture Center building and Perror PGE Gas P Gas Engine P GE GE-SP Engine BEMS. Controller Plant

Control System of PNM Figure 12 Coordinated Control of gas engine and battery Photovoltaic Battery 500kW 2MW The result of the demonstration with Coordinated Control is shown

Facilities Installed By PNM in Figure 13. The experiment demonstrated that, in this case, the capacity of storage battery required to perform PV output smoothing Building smart grid can be lowered by approximately 14% with Coordinated Control

BEMS compared to the case without Coordinated Control.

PI Server However, it must be noted that there factors such as FC GE 60+20kW 180±60kW communication delay can affect control performance. In the verification -30kW Facilities Installed By NEDO based on actual data (see Figure 14), it was indicated that the Figure 10 BEMS2 Configuration – Interconnected Operation- Coordinated Control has the potential to reduce battery capacity by up to 17.7% with a lower communication delay. Commercial Building Open CB BEMS 10 8 6 PV FC GE BAT 4 Load 2 50kW 80kW 200kW ±100kW kWh 0 -2 1 1201 2401 3601 4801 6001 7201 8401 9601 10801 12001 13201 -4 -6 Figure 11 BEMS2 Configuration – Interconnected Operation- Time(s) Battery capacity for PV Battery capacity with GE+FC for PV For grid-connected mode, the BEMS2 of Tokyo Gas was Figure 14 Verification of battery capacity reducing effect based on controlled to operate in conjunction with the utility-owned PV plant and actual data energy storage system. In Shimizu’s BEMS1, on the other hand, operated in conjunction with the small-scale PV and small-scale energy storage within the microgrid during grid-connected mode.

- 6 -

NEDO Microgrid Case Study

Figure 13 Performance of Coordinated Control

For the islanding mode, devices to be controlled are basically the The results of experimental demonstration with BEMS2 in same as in the case of Shimizu BEMS1, but the concept of control is islanding operation mode are shown in Figures 16. It is indicated that different. As shown in the control logic in Figure 15, in the Tokyo Gas the power supply remained stable because of the distributed energy BEMS2, the gas engine was considered as the main energy resource resources such as gas engine even when the load changed drastically. within the island. In other words, it is basically the gas engine that takes charge of absorbing load fluctuations when in the islanding mode. 340 300 260 200kW 220 LOAD Load Load=P +P +P GE FC BT 180 ZONE1 PFCW=30kW YES 140 ST W Load＜110 PV Off command

Power(kW) 100 PBTW=Load-PFC-120 FC NO GE 120kW 90kW + 60 ZONE2 YES PFCW=30kW Ｂ 110≦Load＜160 20

PBTW=Load-PFC-120 NO -20 1 301 601 901 1201 1501 1801 2101 2401 2701 + Ｂ -60 Battery ZONE3 YES PFCW=Load-120 PGE = 120kW 160≦Load＜200 -100 (Prospect) Time(s)

PBTW= Load-120-PFC NO BAT output BAT set GE output FC output FC set LOAD PV + Ａ Ｂ Figure 16 Results of islanding operation of Tokyo Gas’ BEMS2 +

ZONE4 YES PFCW=80kW 120≦PGE＜ 4.2 Key Findings – Lessons Learned 200≦Load＜280 200kW (Prospect) PBTW＝80－P NO FC Three important lessons can be drawn from the BEMS2 + Ａ Ｂ demonstrations, as described below: STPVW On command +

ZONE5 YES PFCW=80kW 280≦Load＜330 Lesson (1): Knowledge obtained from collaborative research NO PBTW=Load-PFC-200 The approach based on BEMS2 which Tokyo Gas developed in + Ａ ZONE6 YES PFCW=80kW cooperation with Sandia National Laboratories and PNM allowed 330≦Load PBTW=STBTup coordination between the microgrid of the building in Mesa Del Sol and Figure 15 Logic of islanding operation of Tokyo Gas’ BEMS2 PNM’s large scale PV and storage system. This was the first - 7 -

NEDO Microgrid Case Study experience for microgrid community to integrate and operate the found only after we actually implemented the demonstration respective system of Japan and the U.S. Dr. Abraham Elis of Sandia experiment. The experiments produced the expected results, but also National Laboratories spoke about the significance of this experience produced new lessons learned for all of us. below. Collaborative efforts in the project deepened mutual understanding of technology challenges, and promoted a shared vision Lesson (2): Demonstrated performance of gas engine to contribute for future developments. This type of demonstration involving actual to the grid operation hardware and collaboration among customers, equipment Through this BEMS2-based approach which Tokyo Gas manufacturers, researchers and utilities greatly contributes to the future developed in cooperation with Sandia National Laboratories and PNM, development of smart grid. it was verified that CHP systems such as a gas engine and fuel cell are highly effective for the local compensation of variable energy sources Comment by Dr. Abraham Ellis including PV. By operating DERs at partial load, there will be a margin The collaborative research with the stakeholders in Japan and in New for the adjustment. If we can control this margin in an appropriate Mexico has been very important for Sandia National Laboratories and manner, we can reduce the control actions that the power system is contributed to a very successful project. required to perform, thereby contributing to the proper operation of The fact that various stakeholders in Japan and the U.S. conducted a power system. Dr. Takao Shinji of Tokyo Gas, who led the project, research collaboratively, shared a vision and discovered areas where described his experience with the experiment as follows: they could work together was an important contribution to the future cooperation between the two countries in this field. Comment by Dr. Takao Shinji Microgrid is the technology that facilitates modernization of electric I am satisfied with this experimental demonstration because the system by enabling efficiency improvement, maximum use of renewable expected results have been verified. energy and resilient energy supply. It has an important implication in terms of the proven universality of We would like to further promote the research based on the fruitful this solution that it was demonstrated here in New Mexico with the results of this research project. assistance of Sandia National Laboratories and PNM that gas engine, fuel cell and other power generation devices for customers can It was demonstrated in this collaborative research that a number of contribute to stabilizing power system without losing the original distributed generators in different sites could be controlled for a single function as power-generating equipment. common purpose. The demonstration experiment was successfully completed in this regard, but there were also challenges that surfaced In general, energy storage as a solution for high penetration of PVs only after carrying out the experiment. Dr. Abraham Ellis from Sandia will be introduced in order to stabilize power system as a new function. National Laboratories reflected on an important lesson learned for future On the other hand, this demonstration shows that gas engines and fuel consideration on communication-related requirements. cells that have been already introduced by customer could also be used for the stabilization of power system. If these existing CHP systems can Comment by Dr. Abraham Ellis be used not only as the original power-generating facility but also as the The Prosperity Site (PNM’s large PV-storage demonstration) and the contributing factor to the power system operation, the effectiveness of Aperture Center in Mesa Del Sol are physically separated, so the key DERs can be further enhanced. issue was the requirement for remote communication. We were also It should be noted here that the generation efficiency of gas engine interested in whether the electric utility can control the distributed and fuel cell would drop if they contribute to power grid operation. If generations on demand side. The two systems were not designed to there is a substantial decline in the generation efficiency because of the interoperate, and enabling full communication for control purposes was partial load operation, the contribution to power grid operation may be a challenge. limited by cost considerations. Figure 17 shows the characteristics of We have to consider how the communication interfaces should be partial load operation of gas engine and fuel cell. In partial load standardized in the future. operation, gas engines have the characteristic that generation efficiency decreases but heat recovery efficiency increases. For fuel cell, its heat Although the simulated reduction in storage battery capacity was recovery efficiency drops but the generation efficiency improves. That 17.7% as previously described, the performance of actually built system is to say, partial load operation may not have a significant impact on the resulted in less than 17.7% due to communication delay, etc. As Prof. overall efficiency of an entire system. From these features, there is Andrea Albert Mammoli commented, there are many factors that we ample room for us to study on the contribution of gas engines and fuel

- 8 -

NEDO Microgrid Case Study cells to power system operation so as to realize the overall optimization 5. Acknowledgment of power system operation. The author gratefully acknowledges the contributions of the following people to this work: Mr. Atsushi Denda and Dr. Kimio Lesson (3): Possible islanding operation mainly by gas engine Morino (Shimizu Corporation), Dr. Takao Shinji and Mr. Masayuki Arguments on islanding operation of microgrid often focus on Tadokoro (Tokyo Gas), Dr. Abraham Ellis and Mr. Jay Johnson (Sandia energy storage including batteries because of their excellent National Laboratories), Mr. Jonathan Hawkins (PNM Resources), Prof. responsiveness. For BEMS1 of Shimizu, the method of islanding Andrea Alberto Mammoli (University of New Mexico) and Mr. operation has been established to fully utilize the batteries. Manuel Barrera (Mesa Del Sol). For BEMS2 of Tokyo Gas, on the other hand, the method of islanding This Case Study was commissioned by the New Energy and operation has been established based on gas engine instead of utilizing Industrial Technology Development Organization (NEDO). the high capability of energy storage. Compared to batteries, gas engine’s responsiveness is not quick (measured by the ramp rate of the 6. Reference DER). However, the demonstration presented the possibility using a gas [1] NEDO, “Report on Japan-US Collaborative Smart Grid engine for power management of microgrid in islanding state. Demonstration Project, Smart Building Demonstration in Albuquerque,” NEDO FY 2010 – FY 2013 Project Report, 2014.

(in Japanese)

45.0% 45.0% [2] Use Case- “BEMS Control of DERs and HVAC Equipment in a 43.0% 43.0% Commercial Building Which Enables Islanding Operation and Heat recovery efficiency 41.0% 41.0% Demand Response”, posted on the EPRI Smart Grid Use Case Repository. June, 17, 2011. 39.0% 39.0%

37.0% Generating efficiency 37.0% [3] Atsushi Denda; Outline of the Albuquerque smart grid project for

35.0% 35.0% efficiency recovery Heat commercial buildings -Japan-US collaborative smart grid

Generating efficiency Generating 33.0% 33.0% project in New Mexico- , IRED2014, Nov. 18. 2014. 50.0% 60.0% 70.0% 80.0% 90.0% 100.0% Output [4] Kimio Morino; Demonstration of the Smart Grid at a Commercial Building in City of Albuquerque, IRED2014, Nov. (a) Gas Engine 18. 2014.

46.0% 23% [5] J. Johnson, K. Morino, A. Denda, J. Hawkins, B. Arellano, T. Ogata, T. Shinji, M. Tadokoro, A. Ellis, “Experimental 45.5% 21% Comparison of PV-Smoothing Controllers using Distributed 45.0% 19% Generating efficiency Generators”, Sandia Report SAND2014-1546, 2014 44.5% 17% [6] A. Ellis and D. Schoenwald, “PV Output Smoothing with Energy 44.0% 15% Heat recovery efficiency Storage,” Proceedings of the 38th IEEE Photovoltaic Specialists 43.5% 13% efficiency recovery Heat Conference, Austin, TX, 3–8 June 8, 2012. Generating efficiency Generating 43.0% 11% 50.0% 60.0% 70.0% 80.0% 90.0% 100.0% [7] J. Johnson, A. Ellis, A. Denda, K. Morino, T. Shinji, T. Ogata, M. Output Tadokoro, “PV Output Smoothing using a Battery and Natural

Gas Engine-Generator,” 39th IEEE Photovoltaic Specialists (b) Fuel Cell Conference, Tampa Bay, Florida, 16-21 Jun, 2013. Figure 17 Characteristics of the partial load operations of gas engine [8] Interviews with stakeholders by author and fuel cell

When designing a microgrid, whether to have an integrated system including energy storage or to make a simple CHP-based system largely depends on the decision made by the entity who adopts the equipment.

However, as mentioned above, the two experimental demonstrations conducted in Albuquerque showcase two viable technical solutions.

© 2015 New Energy and Industrial Technology Development Organization. All rights reserved. - 9 -