HeatSource Propane Case Study

The Challenge

Propane delivery in northern New England is not a convenience. It is survival. When temperatures drop to -10°F and a customer’s tank runs empty, the consequences are frozen pipes, damaged property, and in the worst cases, a genuine threat to health and safety. Tom Garrity understood this better than most. As dispatch supervisor for HeatSource Propane, he managed deliveries to more than 3,000 residential and commercial accounts spread across some of the most remote terrain in the northeastern United States.

For years, HeatSource ran deliveries the way most rural propane companies did: by territory and driver memory. Each of the 15 drivers was assigned a geographic zone they knew well. They drove the same roads to the same customers on schedules based largely on habit and rough estimates of when tanks might need refilling. Dispatchers called customers to confirm delivery windows. During normal weather, the system worked well enough.

Cold snaps changed everything.

When forecasts dropped below zero for extended periods, propane consumption spiked. Customers who normally went three weeks between fills burned through their tanks in ten days. The office phone lines became overwhelmed with emergency run-out calls. Dispatchers answered calls back-to-back for hours, trying to triage which customers were most urgent while drivers followed pre-set routes that didn’t account for the sudden shift in demand.

The specific problems were critical:

• 24 to 48-hour emergency response times: When a customer called reporting an empty tank in sub-zero weather, the earliest HeatSource could guarantee a fill was the next day. During extended cold snaps, it stretched to two days. Dispatchers had no way to identify which truck was nearest to the emergency or had capacity remaining.
• No usage-based scheduling: Delivery schedules relied on static intervals and driver memory. A customer who used propane for heating, cooking, and hot water received the same delivery cadence as one who used it only for a backup generator. Heavy-use customers ran out. Light-use customers received deliveries they didn’t need yet.
• Office staff buried in phone coordination: Two dispatchers spent their entire workday calling customers to schedule deliveries, fielding run-out calls, and relaying instructions to drivers. During peak demand, the phones didn’t stop ringing from 6am to 6pm.
• The incident that changed everything: During a January cold snap, an elderly customer in a remote area of Vermont ran out of propane and went without heat for 36 hours in -10°F conditions. A neighbor discovered the situation and called local news. HeatSource avoided a full PR crisis only because the neighbor had brought the customer to their home. But the incident made clear that the old system was not just inefficient. It was dangerous.

The Solution

Tom began evaluating route optimization platforms the week after the incident. Most tools he looked at were designed for urban delivery fleets running dozens of short stops per day. HeatSource needed something different: a platform that could handle rural routes spanning 80+ miles of back roads, optimize across 15 trucks with varying tank capacities, and incorporate usage-based delivery scheduling so customers received propane before they ran out, not after.

He found Upper through a recommendation from another fuel delivery company in Maine. The platform’s ability to import large customer lists with custom delivery dates, optimize multi-driver routes on rural road networks, and provide live GPS tracking checked every requirement on his list.

Usage-Based Scheduling Replaces Guesswork

The first step was replacing the territory-and-memory system with data-driven delivery scheduling. Tom’s team exported their customer database, including tank sizes, average usage rates, and last fill dates, into a CSV file. They imported the full list into Upper with calculated next-delivery dates based on each customer’s consumption pattern.

Heavy-use residential customers who burned through 100 gallons per week in winter were scheduled on tighter intervals than commercial accounts with large storage tanks. The system generated optimized delivery scheduling and grouped nearby customers due for fills on the same day, rather than sending drivers out to the same road three times in a week for three different customers.

The scheduling shift produced immediate results. Run-out calls dropped because customers received propane before their tanks hit empty. Drivers made fewer trips because stops were clustered by geography and timing. The dispatchers who had spent their days calling customers to confirm schedules now had the delivery calendar managed through the platform.

Emergency Dispatch Through Nearest-Truck Routing

The second transformation addressed the emergency response problem that had nearly become a crisis. Before Upper, an emergency run-out call triggered a chain of phone calls. The dispatcher called drivers one by one to find someone who could make a detour, then gave verbal directions. Response time depended on which driver happened to answer the phone first, not which one was actually closest.

With Upper’s live GPS tracking, Tom could see the real-time position of every truck on a single map. When an emergency call came in, he identified the nearest truck with remaining capacity, added the stop to that driver’s active route, and the driver received the updated route on their phone within seconds. No phone calls. No guessing. No 24-hour wait.

During the following winter, HeatSource’s emergency response time dropped from a 24-48 hour average to under 4 hours. In several cases, Tom dispatched a truck within 45 minutes of receiving the call because the nearest driver was already on a route less than 10 miles away.

The customer notification system added a layer of communication that had never existed before. When a delivery was en route, customers received automated updates with an estimated arrival window. For elderly customers and those in remote locations, this was particularly valuable. They knew when to expect their delivery without needing to call the office.

DOT Compliance and Route Documentation

Propane delivery trucks are subject to Department of Transportation regulations that require documented route records. Before Upper, drivers kept paper logs that were incomplete and difficult to audit. Upper’s breadcrumb trail feature records every route driven with GPS coordinates, timestamps, and stop-by-stop delivery records. When a DOT auditor requested route documentation, Tom could pull complete records for any driver on any date.

The vehicle tank capacity configuration also improved route accuracy. Each of HeatSource’s 15 trucks had different tank sizes. Upper’s optimization factored in how many gallons each truck carried, ensuring drivers weren’t assigned more stops than their truck could service. This eliminated the mid-route refueling trips that had been adding 30-45 minutes to some drivers’ days.

The Impact

The transformation at HeatSource Propane was measured in numbers, but felt in the quiet absence of crisis calls. The winter following the implementation was one of the coldest on record in Vermont. Temperatures dropped below zero for 18 consecutive days in February. In previous years, that stretch would have generated hundreds of emergency calls, overwhelmed the dispatch team, and left vulnerable customers waiting days for fuel. Instead, the usage-based scheduling anticipated the surge. Drivers ran optimized routes that pre-filled customers before tanks dropped below 20%. Emergency calls during that cold snap totaled a fraction of the previous year.

Customer churn, which had been running at 5% annually, dropped to 1.5%. The improvement came from two factors: fewer run-outs meant fewer angry customers, and the delivery notifications gave customers confidence that HeatSource was proactively managing their account. Several customers who had been considering switching to a competitor cited the tracking notifications as their reason for staying.

The fleet efficiency gains translated directly to fuel savings. Optimized routes reduced total miles driven by 15%, saving $6,200 per month in diesel costs across the 15-truck fleet. Drivers completed their routes faster, which reduced overtime during peak season and gave the team capacity to take on new customers without adding trucks.

Most importantly, there have been zero elderly-customer heat incidents since implementation. Tom set up priority flags for customers over 75 and those in remote locations with no nearby neighbors. These accounts are scheduled on tighter delivery intervals and flagged for immediate dispatch if a run-out call comes in.

The dispatchers’ workday changed entirely. Instead of spending eight hours on phone calls coordinating schedules and relaying emergency instructions, they manage the delivery calendar through Upper’s dashboard and handle exceptions. The proactive scheduling means fewer customers call in because their deliveries arrive before they need to ask.

Performance Metrics

Tom still thinks about the Mrs. Pelletier incident. He keeps a photo of the temperature readout from that January day pinned above his desk. But it serves as a reminder of a problem that no longer exists.