Erlang Calculator

What is Erlang Calculator?

An Erlang Calculator is a specialized analytical tool used to model telephone traffic, call center staffing, and telecommunications network capacity. Named after the Danish mathematician Agner Krarup Erlang, this calculator applies queuing theory to determine how many lines or agents are needed to handle a given volume of calls at a specific service level. It is the backbone of modern contact center workforce management and network engineering, converting raw call data into actionable staffing requirements.

Call center managers, telecommunications engineers, and operations analysts rely on the Erlang Calculator to balance service quality against operational costs. By inputting metrics like call arrival rate, average handling time, and desired wait time, users can forecast the exact number of trunks or agents required to meet service level agreements (SLAs). This prevents both overstaffing (which wastes money) and understaffing (which leads to long hold times and customer frustration).

This free online Erlang Calculator provides instant results without requiring complex spreadsheet formulas or expensive workforce management software. It supports both Erlang B (for trunk lines) and Erlang C (for queued call centers) models, making it versatile for network design and contact center planning alike.

How to Use This Erlang Calculator

Using this Erlang Calculator is straightforward, even if you have no background in queuing theory. Simply enter your call traffic data into the designated fields, select the appropriate model, and the calculator will instantly compute the required resources. Follow these five steps to get accurate results every time.

1. Select the Erlang Model: Choose between Erlang B (for loss systems like telephone trunks) or Erlang C (for delay systems like call center queues). Erlang B assumes calls are blocked and lost if no line is available, while Erlang C assumes calls wait in a queue. Your choice depends on whether your system allows queuing.
2. Enter Traffic Intensity (Erlangs): Input the total traffic load in Erlangs, which represents the number of simultaneous calls during the busiest hour. Calculate this by multiplying the number of calls per hour by the average call duration in hours. For example, 100 calls per hour with an average handling time of 6 minutes equals 10 Erlangs (100 × 0.1 hours).
3. Set the Blocking Probability (Erlang B) or Service Level Target (Erlang C): For Erlang B, enter the acceptable blocking probability (e.g., 0.01 for 1% blocked calls). For Erlang C, enter your target service levelΓÇötypically 80% of calls answered within 20 seconds. This defines the quality of service you aim to achieve.
4. Enter Average Call Duration (Optional for Erlang C): If using Erlang C, input the average handling time (AHT) in seconds or minutes. This parameter directly affects queue wait times and agent occupancy. Shorter calls require fewer agents, while longer calls increase staffing needs.
5. Click Calculate: Press the "Calculate" button to generate results. The output will display the minimum number of lines (Erlang B) or agents (Erlang C) needed, along with key performance indicators like expected wait time, agent occupancy, and probability of delay.

For best accuracy, always use data from your busiest 30-minute or 60-minute period. Avoid averaging traffic across an entire day, as peak hour spikes drive staffing requirements. If you are unsure about traffic intensity, most call center software can export this metric directly from historical reports.

Formula and Calculation Method

The Erlang Calculator uses two primary mathematical models: Erlang B and Erlang C. Erlang B is derived from the Poisson distribution and assumes a finite number of servers with no queuing (blocked calls are lost). Erlang C extends this by assuming an infinite queue where calls wait until an agent becomes available. Both formulas are rooted in Markov chain theory and are widely accepted in telecommunications engineering.

In these formulas, A represents the total traffic intensity in Erlangs (arrival rate × average service time), and N is the number of servers (lines or agents). The Erlang B result P_B is the probability that all N lines are busy and a call is blocked. The Erlang C result P_W is the probability that a call must wait in the queue. Both formulas rely on factorial calculations and summations, which become computationally intensive for large N, making an online calculator essential for practical use.

Understanding the Variables

Traffic Intensity (A) in Erlangs: This is the fundamental input, defined as the product of call arrival rate (λ, calls per unit time) and average service time (h, in same time units). One Erlang equals one hour of call traffic per hour, or 60 call-minutes per hour. For example, 30 calls per hour with 2-minute average duration = 30 × (2/60) = 1 Erlang. This dimensionless unit directly represents the average number of simultaneous calls in progress.

Number of Servers (N): In Erlang B, N is the number of trunk lines or channels. In Erlang C, N is the number of agents or customer service representatives. The calculator finds the minimum N that satisfies your blocking or service level target. Increasing N reduces blocking probability but increases cost.

Blocking Probability (P_B): For Erlang B, this is the fraction of calls that receive a busy signal or are rejected. A typical target is 0.01 (1%) for voice networks, though data networks may tolerate higher rates. Lower blocking requires more lines.

Service Level Target (SL): For Erlang C, this defines the percentage of calls answered within a threshold time (e.g., 80% in 20 seconds). The calculator uses this to compute required agents, factoring in queue wait times and agent occupancy.

Step-by-Step Calculation

For Erlang B, the calculation begins by computing the factorial of N (N!) and the power A^N. The denominator sums all terms A^k/k! from k=0 to N. The numerator is A^N/N!. The blocking probability is the ratio. Since factorials grow rapidly, the calculator uses iterative algorithms like the Erlang B recursion formula: P_B(A, N) = (A × P_B(A, N-1)) / (N + A × P_B(A, N-1)), starting with P_B(A, 0) = 1. This avoids overflow and speeds computation.

For Erlang C, the calculator first computes the probability of delay (P_W) using the extended formula. Then, it applies the M/M/N queue model to estimate average wait time for delayed calls: E[W|delay] = (A^N / (N! × N × μ × (1 - A/N)²)) × P_W, where μ = 1/h. The service level is then derived from the exponential distribution of wait times. The calculator iterates N until the target service level is met, making it a root-finding problem solved efficiently by binary search.

Example Calculation

Let us walk through a realistic scenario for a mid-sized customer support call center using the Erlang C model. This example demonstrates how to translate business requirements into staffing numbers using the calculator.

Step 1: Calculate Traffic Intensity (A). Convert AHT to hours: 7 minutes = 7/60 = 0.1167 hours. Multiply by call arrival rate: A = 450 × 0.1167 = 52.5 Erlangs. This means at any given moment, there are an average of 52.5 simultaneous calls in progress.

Step 2: Estimate Initial Agent Count. A rough rule of thumb is to start with N = A + (0.5 to 1.0) × √A. For A=52.5, √A ≈ 7.25, so N ≈ 52.5 + 7.25 = 60 agents. Enter N=60 into the Erlang C calculator with A=52.5 and target service level 80/20.

Step 3: Compute Results. The calculator shows that with 60 agents, the probability of delay (P_W) is 0.45 (45% of calls wait). The average wait for delayed calls is 12 seconds. The overall service level is 72% answered within 20 seconds, which is below the 80% target. The calculator then iterates to N=63 agents.

Step 4: Final Outcome. With 63 agents, P_W drops to 0.32, average wait for delayed calls is 8 seconds, and service level reaches 81.5% within 20 seconds. Agent occupancy is 83.3%, which is healthy (target 75-85%). The result means the center needs 63 agents to meet the 80/20 service level, not the initial 60.

In plain English, failing to staff these extra three agents would result in 8% more callers waiting longer than 20 seconds, potentially increasing customer frustration and abandonment rates. This calculation justifies the additional headcount to management.

Another Example

Consider a telecommunications company designing a trunk group for a new office building using the Erlang B model. The building expects 200 calls per hour during peak, with an average call duration of 3 minutes. The acceptable blocking probability is 0.5% (0.005). Traffic intensity A = 200 × (3/60) = 10 Erlangs. Using the Erlang B calculator, with A=10 and target P_B=0.005, the result shows N=17 lines are required. With 16 lines, blocking rises to 0.008 (0.8%), exceeding the target. This tells the network engineer to provision 17 voice channels to ensure fewer than 1 in 200 calls receives a busy signal.

Benefits of Using Erlang Calculator

An Erlang Calculator transforms abstract call traffic data into precise, cost-saving decisions. Whether you manage a small help desk or a large network operations center, this tool delivers quantifiable advantages that directly impact your bottom line and customer experience. Below are the key benefits you gain by incorporating Erlang calculations into your planning process.

• Optimized Staffing Costs: By calculating the exact number of agents needed to meet service level targets, you eliminate guesswork and avoid overstaffing. A typical call center can reduce labor costs by 10-20% by using Erlang C instead of simple ratio-based staffing. For a center with 100 agents, this translates to annual savings of $200,000 or more in wages and benefits.
• Improved Customer Experience: Erlang modeling ensures that enough resources are available during peak hours to keep wait times low. Customers experience shorter hold times, fewer abandoned calls, and higher first-call resolution rates. Studies show that reducing average speed of answer by 10 seconds can increase customer satisfaction scores by 5-8 points.
• Accurate Network Capacity Planning: For telecom engineers, Erlang B prevents overprovisioning of expensive trunk lines while maintaining Grade of Service (GoS). Each unnecessary T1 or E1 line costs thousands per year. The calculator identifies the minimum number of circuits required, freeing capital for other infrastructure investments.
• Data-Driven Decision Making: The calculator provides objective metricsΓÇöblocking probability, agent occupancy, probability of delayΓÇöthat can be presented to stakeholders with confidence. This replaces anecdotal "we need more staff" arguments with hard numbers, making budget approvals easier and more frequent.
• Scalability and Forecasting: By adjusting input parameters, you can model growth scenarios. For instance, if call volume increases by 15% next year, the calculator immediately shows the new staffing requirement. This enables proactive hiring and training schedules, preventing last-minute scrambles during seasonal peaks.

Tips and Tricks for Best Results

To extract maximum value from the Erlang Calculator, you need to apply industry best practices and avoid common pitfalls. These tips come from decades of workforce management experience and can dramatically improve the accuracy of your results.

Pro Tips

• Always use data from your busiest 30-minute interval, not the full hour average. Call arrival rates are rarely uniform; a 30-minute peak can be 30% higher than the hourly average. Using the hourly average will understaff your peak.
• Include shrinkage factors in your final agent count. Shrinkage accounts for breaks, training, meetings, and absenteeism. Multiply the raw Erlang C result by 1/(1 - shrinkage%). For example, with 15% shrinkage, multiply by 1.176 to get the actual number of body hires.
• For Erlang C, set the service level target realistically. An 80/20 target (80% in 20 seconds) is standard for inbound call centers. Pushing to 90/20 requires 15-25% more agents, which may not be cost-justified unless your business demands premium service.
• Validate your Erlang B results with the inverse calculation. If you have N lines and want to know the maximum traffic they can handle at a given blocking probability, use the calculator's inverse mode. This helps in capacity planning for existing infrastructure.

Common Mistakes to Avoid

• Using Average Call Duration Instead of Handling Time: Do not confuse average call duration with average handling time (AHT). AHT includes talk time plus after-call work (ACW). Using only talk time underestimates agent occupancy and leads to understaffing. Always include ACW in your input.
• Ignoring Call Abandonment: Erlang C assumes calls wait indefinitely, but in reality, callers hang up. If your center has high abandonment (over 5%), the Erlang C results will overstate required agents. Use an adjusted service level or consider Erlang A (which models abandonment) for more accuracy.
• Applying Erlang B to Queued Systems: Erlang B assumes blocked calls are lost forever. If your system allows queuing (e.g., an automated attendant that holds calls), use Erlang C instead. Mixing models leads to grossly inaccurate trunk or agent counts.
• Rounding Agent Counts Down: The calculator often yields non-integer results like 47.3 agents. Rounding down to 47 creates a service level deficit. Always round up to the nearest whole number, then apply shrinkage. Fractional agents do not exist in practice.

Conclusion

The Erlang Calculator is an indispensable tool for anyone responsible for telecommunications network design, contact center staffing, or service level management. By applying the mathematically proven Erlang B and Erlang C models, you can precisely determine the number of trunk lines or customer service agents required to handle your traffic load while meeting quality targets. This eliminates wasteful overprovisioning and prevents the customer dissatisfaction that comes from understaffing. Understanding the variablesΓÇötraffic intensity, blocking probability, and service levelΓÇöempowers you to make decisions backed by data rather than intuition.

We encourage you to use this free Erlang Calculator for your next staffing review or network capacity project. Enter your real-world call data, experiment with different service level targets, and see how small changes in traffic intensity impact resource requirements. Bookmark this page for quick access during quarterly planning sessions, and share it with your workforce management team. Accurate resource planning starts hereΓÇöcalculate now and optimize your operations today.

Frequently Asked Questions