Optimal Demand Response-Based Energy Management for Campus Buildings with Battery Energy Storage Systems: A Case Study of the ELB Building

1. Introduction

• DR flexibility is limited by load operating requirements, while BESS can shift energy and reduce peak demand; their economic value depends on load profile, resource configuration, and tariff.A BESS can charge during low-cost or PV-available periods and discharge during high-cost or high-demand periods.Source: §1 · p. 2
• Tariff-oriented studies often use generalized tariff classes or collections of building profiles, motivating evaluation with a facility’s actual tariff and measured operating data.The paper notes that DR and BESS value depends on the billing mechanism.Source: §1 · p. 2
• Campus buildings combine loads with different flexibility, motivating assessment tied to an institutional building’s measured load and applied tariff.Prior campus studies examined price-based DR, PV, storage, and energy management; real-building studies have used measured smart-meter data.Source: §1 · p. 2
• The paper formulates a MILP that jointly schedules flexible demand and BESS under operating constraints, using measured ELB demand and rooftop-PV data and the actual DTE tariff.It compares baseline, DR-only, BESS-only, and coordinated DR–BESS cases.Source: §1 · pp. 2–3
• 25.8%: coordinated DR and BESS reduced the May 2026 monthly bill, compared with 12.0% for DR alone.The study also examines additional BESS capacities and a December 2025 case.Source: §1 · p. 3

2. System Configuration and Operation Strategy

• Figure 1 separates the framework into an optimization layer and a physical layer, with the grid, rooftop solar, building bus, and proposed BESS in the physical system.The BESS is modeled as a proposed addition, not existing equipment.Source: §2 · p. 3, Figure 1 · p. 3
• Figure 2’s first stage prepares hourly billing-period inputs by separating building load into fixed and flexible HVAC and lighting components and defining DR limits and operating constraints.Measured load, solar generation, and utility-tariff data are prepared through the final hour of the billing period.Source: §2 · p. 4, Figure 2 · p. 4
• The second stage minimizes monthly energy, demand, and fixed service charges while scheduling flexible load and BESS charging and discharging, yielding grid import, PV use, billed peak, and total cost.The optimization is subject to DR, PV-use, BESS-operation, and power-balance constraints.Source: §2 · p. 4

3. Problem Formulation

• The objective minimizes grid-energy, BESS degradation, monthly peak-demand, and fixed service charges, with demand-response savings reflected through lower grid imports.Load shedding is not a separate cost term.Source: §3 · pp. 4–5
• Demand response includes load shedding limited by flexible-load availability and an event window, plus load shifting to another hour.HVAC and lighting determine the maximum sheddable load.Source: §3 · p. 5
• Load shifting is constrained by available flexibility and receiving-hour headroom, and total shifted-out demand must equal total shifted-in demand over the cycle.This models shifting as relocation rather than elimination of demand.Source: §3 · p. 5
• BESS constraints cap charging and discharging, prohibit simultaneous operation, track stored energy between hours, and require the ending charge to match the starting charge.Stored energy is also kept within physical limits.Source: §3 · pp. 5–6
• The model limits PV use to available generation, bounds grid import by service capacity, defines billed peak from on-peak imports, and enforces interval power balance.The billing cycle spans a month, aligning the schedule with the tariff's billing period.Source: §3 · pp. 4, 6

• The ELB case study is a 123,000-square-foot facility with engineering laboratories and classrooms, supplied by dedicated electrical metering.The building is billed under the DTE tariff structure.Source: §4 · p. 6
• Multi-year 15-minute load readings were aggregated to hourly averages, and the study tests May 2026 and December 2025, each with 744 hourly data points.May represents the higher-demand cooling season and December the lower-demand winter season; hourly solar readings were available from September 2025.Source: §4 · p. 6
• The DTE tariff combines energy time charges with a demand charge based on the customer’s peak power draw during on-peak hours.The energy and demand charges are summarized in the tariff table.Source: §4 · p. 7, Table 3 · p. 7
• A short-duration peak reduction can yield a cost benefit comparable to a much larger reduction in total energy use because demand charges apply to a single hour.The paper identifies this asymmetry as motivation for its demand-charge-aware objective.Source: §4 · p. 8

5. Simulation Results and Discussion

• The MILP simulations compare baseline, DR-only, BESS-only, and combined DR+BESS operation over May and December billing periods.The primary battery case used 200 kWh usable capacity; additional 100 kWh and 150 kWh cases examined capacity effects.Source: §5 · p. 8
• The model assumes a 20 kWh minimum reserve and allows HVAC reductions up to 30% and lighting reductions up to 15% during DR events.The 200 kWh battery has 92% charging and discharging efficiencies, with its initial and final charge set to half capacity.Source: §5 · p. 8
• The baseline monthly bill was $5,471, with demand charges accounting for about $2,889, or 53% of the bill.The paper concludes that peak management, rather than energy conservation, is the dominant cost saver in this case.Source: §5 · p. 8
• At 200 kWh, BESS-only savings were 9.7%, compared with 12.0% for demand response; the paper identifies demand response as the primary cost-saving measure and storage as complementary.Across the battery scenarios, the reported savings came from reducing demand charges.Source: §5 · p. 9, Table 5 · p. 9
• 25.8% in May and 22.0% in December: coordinated DR and BESS reduced the monthly bill by these amounts, respectively.The December reduction is lower; the paper attributes the difference to December’s flatter load profile and lower baseline peak.Source: Table 5 · p. 9, Table 8 · p. 9, §5 · p. 10

• The study presents a MILP-based framework for scheduling demand response and battery storage under an accurate commercial electricity tariff.Source: §6 · p. 10
• 25.8% bill reduction and a billed-peak drop from 134.5 kW to 76.4 kW came from coordinated demand response and battery storage.Source: §6 · p. 11
• Demand response alone reduced the monthly bill by 12.0%, outperforming battery alone at 9.7%, while the two measures were complementary.Source: §6 · p. 11
• Across the four cases, energy charge varied by less than 7% while demand charge fell by $1,248, locating savings almost entirely in the demand charge.Source: §6 · p. 11
• The authors identify no-investment demand response as a logical first step for institutional buildings, with storage deployed as a complement.Source: §6 · p. 11
• Future work will evaluate a full year and determine the optimal battery and payback period.Source: §6 · p. 11