Sample and velocity traverses

EPA Method 1, 40 CFR Part 60, Appendix A-1

Before any gas is sampled, Method 1 decides where on the stack to measure and exactly which points inside it to visit. Every number later in the test inherits these choices.

Answers: Where do the ports go, how many points are needed, where is each point, and is the flow straight enough to measure?
Equipment: Tape or stack gauge, Type S pitot tube, inclined manometer, angle-measuring device
Produces: A traverse point layout and the probe markings for each point
Feeds into: Methods 2, 3, 4, 5 and the instrumental methods

Why one point is never enough

Gas does not move up a stack as a uniform column. Friction slows it near the walls, so the center runs faster than the edges. Anything that bends, narrows, widens, or stirs the flow—an elbow, a fan, a damper, a scrubber inlet—pushes the fast-moving core off to one side, and it takes several diameters of straight duct for the flow to settle again.

Pollutant concentration can be uneven too, especially for particulate, where heavy particles get flung to the outside of bends. A single measurement at one spot can therefore be badly wrong. Method 1 solves this two ways: it pushes the sampling location far from disturbances, and it spreads measurements over a grid of points that each represent an equal share of the cross-section.

Flow recovering after a bend

Move the measurement plane along the duct.

Distance downstream of the bend

Peak ÷ average velocity

—

Error, one center point

—

Error, 12-point traverse

—

A simplified model: a 1/7-power turbulent profile skewed toward the outer wall of the bend, with the skew decaying downstream. The area-average velocity stays the same at every distance.

Even in perfectly settled flow, the centerline runs about 22% faster than the average. The equal-area traverse lands within about 1% because each point stands for the same slice of area.

Choosing the measurement site

Method 1 measures distances in stack diameters, so the same rule works on a 2-foot vent and a 20-foot power plant stack. Two distances matter:

Distance B runs from the nearest disturbance upstream of the port (where the gas came from) to the port. The flow needs this length to recover.
Distance A runs from the port to the nearest disturbance downstream (where the gas is going), often the stack exit. Flow starts to rearrange itself before it reaches an obstacle.

The ideal site is at least 8 diameters downstream and 2 diameters upstream of any disturbance: the “8-and-2 rule.” A disturbance is anything that changes the flow—bends, expansions, contractions, fans, dampers, control devices, or a visible flame. Meet 8-and-2 and you sample the minimum number of points. Fall short and you need more points, read from Figure 1-1 of the method. Below 2 diameters downstream or 0.5 diameters upstream the site is unacceptable without the regulator’s approval.

Equivalent diameter for a rectangular duct

D_e = \frac{2\,L\,W}{L + W}

L, W: inside length and width of the duct

Site planner and Figure 1-1

Drag the port up and down the stack, or use the sliders.

Stack shape

Inside diameter

Straight run between disturbances

Port height above the inlet

Distance B

—

Distance A

—

Minimum points

—

Figure 1-1 applies to particulate traverses. Read the number of points for distance B (bottom axis) and for distance A (top axis), then use the larger. For a circular stack, round up to a multiple of four so each of the two diameters gets the same count.

Minimum points when the 8-and-2 rule is met

12 points for circular or rectangular stacks wider than 24 in. (0.61 m)
8 points for circular stacks 12 to 24 in. (0.30–0.61 m)
9 points for rectangular stacks with an equivalent diameter of 12 to 24 in.

Method 1 applies to stacks larger than 12 in. diameter (or 113 in.² area). Smaller ducts, from 4 to 12 in., use Method 1A. Velocity-only traverses can use Figure 1-2, which allows fewer points than Figure 1-1.

Locating the points: equal areas

Each traverse point must represent the same share of the cross-sectional area, so that a simple average of the readings is a true area-weighted average. For a circular stack, Method 1 divides the circle into concentric rings of equal area and places each point at the radius that splits its ring into two equal halves. Two traverse diameters, at right angles through two ports, cover the stack.

Because the outer rings are thin and the inner ones are wide, the points crowd toward the walls. That is deliberate: the wall region holds a lot of area even though it looks narrow.

Point location on a diameter, as a percent of the inside diameter from the near wall (Table 1-2)

P_i = 50\left[\,1 - \sqrt{\frac{n - 2i + 1}{n}}\;\right] \qquad i = 1,\dots,\tfrac{n}{2}

n: number of points on one diameter (half the total for a circular stack)
i: point number counted from the near wall
P_{n+1-i}: the far half mirrors the near half: 100 - P_i

Equal-area layout and probe markings

Change the point count and stack size. Shaded bands have equal area.

Stack shape

Total traverse points

Inside diameter

Port depth (nipple + wall)

Keeping points off the wall

On small stacks the outermost points can land so close to the wall that the nozzle or pitot would touch it. Method 1 moves them:

Stacks wider than 24 in.: no point closer than 1.0 in. to the wall, or the nozzle inside diameter if that is larger.
Stacks 24 in. or smaller: no point closer than 0.5 in., or the nozzle inside diameter if larger.

If two neighboring points collapse onto the same adjusted location, you still treat them as two separate points: sample there twice as long and record two readings.

Rectangular ducts

Rectangular ducts are split into a grid of equal rectangles with a point at the center of each. Table 1-1 sets the grid, with more points along the longer side. Each small rectangle should have a length-to-width ratio between 1.0 and 2.0, so the grid stays close to square.

Table 1-1: Grid layouts for rectangular stacks
Points	Matrix layout	Points	Matrix layout
9	3 × 3	30	6 × 5
12	4 × 3	36	6 × 6
16	4 × 4	42	7 × 6
20	5 × 4	49	7 × 7
25	5 × 5

Checking for cyclonic flow

Methods 2 and 5 assume the gas moves straight up the stack. Downstream of cyclones, tangential inlets, wet scrubbers, and some fans, the gas can instead spiral. A pitot tube aimed straight up would then read the wrong velocity, and a nozzle facing straight down would sample at the wrong angle.

The check uses the Type S pitot tube in a sideways position, called the 0° reference: the planes of the two openings are perpendicular to the stack cross-section, so gas moving straight up flows across both openings equally and the manometer reads zero (a “null”). If the gas is swirling, it hits one opening harder and the manometer reads something. You rotate the pitot until the reading returns to zero; the rotation angle \alpha is the flow’s yaw angle at that point.

Null-angle check

Rotate the pitot until the manometer reads 0.00, then record the angle.

Pitot rotation

Swirl in this stack Changing the swirl starts a new check.

Point

—

Manometer

—

Average |α|

—

Points that need no rotation count as α = 0° and stay in the average. If the average of the absolute angles is more than 20°, the site fails: straighten the flow with vanes, move to another location, or use an approved alternative such as the three-dimensional pitot methods (Methods 2F and 2G).

Cyclonic flow criterion

\bar{\alpha} = \frac{1}{N}\sum_{i=1}^{N}\lvert\alpha_i\rvert \;\le\; 20^\circ

Acceptance criteria

Method 1 at a glance

Stack diameter greater than 12 in. (0.30 m), or cross-section greater than 113 in.² (0.071 m²). Otherwise use Method 1A.
Ideal site: at least 8 diameters downstream and 2 diameters upstream of any disturbance.
Absolute minimum: 2 diameters downstream and 0.5 diameters upstream.
Particulate traverses: up to 24 points (circular) or 25 (rectangular) from Figure 1-1.
Points no closer to the wall than 1.0 in. (stacks over 24 in.) or 0.5 in. (24 in. and under), or the nozzle inside diameter if larger.
Rectangular elements with a length-to-width ratio between 1.0 and 2.0.
Average absolute yaw angle no more than 20°.

Field mistakes that cost runs

Measuring from the wrong reference

Point distances in Table 1-2 are from the inside wall, but you mark the probe from the outside face of the port. Forget to add the nipple length and wall thickness and every point shifts toward the far wall.

Forgetting the stack exit

The top of the stack is a disturbance. A port 1.5 diameters below the exit fails the 2-diameter upstream criterion even if the inlet is far below.

Trusting the drawings

Measure the inside diameter at the ports. Stacks get relined, refractory erodes, and some are not round. Check the distance to the far wall through both ports.

Skipping the cyclonic check after a scrubber

Wet scrubbers, cyclones, and tangential inlets are classic sources of swirl. Run the null check before you commit to the location, not after three runs.