FAULT ISOLATION MANUAL · CARBONATE SYSTEM — ALKALINITY

ALK-002 · Alkalinity risingv1.0.1

Alkalinity climbs from test to test while the dosing program sits unchanged. This sheet isolates why, the way an aircraft maintenance manual would: confirm the reading and the trend, clear the confounds, check kalkwasser and dosing inputs, verify delivery, and only then act. Every correction here reduces an input: nothing is dosed to force the level down. Answer each numbered block and the sheet lights the path.

WHAT YOU'LL NEED

Your alkalinity test kit (plus a second kit or fresh reagent)
A graduated container and the dosing calculator
Your kalkwasser setup details, if you run one
Your tank journal

JavaScript is off, so the interactive sheet can't run — the complete printed logic is below. Start at block 1 and follow the YES/NO links. Nothing about your tank is inferred or stored.

Block 1 — Have you confirmed the reading with a second test (a different kit, or fresh reagent)? (Note: Alkalinity reagents age and titration endpoints drift: a mis-read endpoint shifts every reading the same direction.)
YES → block 3 · NO → block 2

Block 2 — Retest before troubleshooting: a single reading is not a rising trend. Confirm with a second kit or fresh reagent, then return to this procedure. (Verify — 2 days, ALKALINITY: A second kit or fresh reagent reproduces the reading within the kits' stated resolution)
END — ESCALATE RETEST

Block 3 — Across at least 3 comparable tests (same kit, same time of day), has alkalinity continued to rise from test to test? (Note: Comparable means same kit, same time of day. Alkalinity swings with the tank's daily cycle, so mixed testing times can manufacture a trend that is not there; and a single high reading is not a rising trend.)
YES → block 9 · NO → block 4

Block 4 — Is alkalinity still inside your tank-type target range: 8–10 dKH (mixed reef) or 7–9 dKH (SPS-dominant)?
YES → block 5 · NO → block 6

Block 5 — ALK-002 NOT ACTIVE — alkalinity is not on a confirmed rising trend and sits inside your tank-type target range. That establishes only that this fault's entry condition is not met. Log the readings in your journal: a recorded baseline makes the next investigation faster.
END — NO FAULT

Block 6 — Is alkalinity ABOVE the range (rather than below it)?
YES → block 7 · NO → block 8

Block 7 — ABOVE TARGET RANGE, TREND NOT CONFIRMED: reduce or suspend your alkalinity inputs and let daily consumption bring the level down into range. Do not dose anything to force it down. Routine water changes with a salt that mixes inside range pull the level toward range. Keep testing on the same comparable schedule; if a rising trend emerges, run this procedure again. (Note: There is no safe additive that lowers alkalinity: consumption and water changes are the tools. Cutting inputs and waiting is the correction.) (Verify — 3 days, ALKALINITY: Alkalinity moving down toward or inside your tank-type target range with inputs reduced across comparable tests)
verified resolved → done · problem continues → block 3

Block 8 — BELOW TARGET RANGE: a low level is not this fault. Restoring a deficit is correction-mode work, and if alkalinity is FALLING from test to test, run ALK-001 (alkalinity decreasing) instead: that sheet isolates the cause before you correct. (Note: ALK-002 isolates a rising trend. Running the wrong sheet wastes the trend window you have already established.) (Verify — 3 days, ALKALINITY: Alkalinity stable or rising toward range across comparable tests, or ALK-001 opened for a confirmed decline)
verified resolved → done · problem continues → block 3

Block 9 — During that SAME trend window: any water change, alkalinity or calcium correction, kalkwasser concentration change, or change to any dosing program? (Note: Replacement water that mixes high in alkalinity, a parallel correction, or a kalkwasser change can mask or mimic a rise: the window has to be clean before the trend is attributed.)
YES → block 10 · NO → block 11

Block 10 — TREND WINDOW CONFOUNDED — the rise cannot be attributed while the window contains an intervention. Re-establish the trend: 3 comparable tests, same kit, same time of day, with no water changes or parallel corrections. If the rise reproduces, return to this procedure. (Verify — 3 days, ALKALINITY: Three comparable tests in an intervention-free window still show alkalinity rising from test to test)
END — ESCALATE RETEST

Block 11 — Is alkalinity currently BELOW your tank-type target range: 8–10 dKH (mixed reef) or 7–9 dKH (SPS-dominant)? (Note: A rise from below range is the level recovering toward target: cutting inputs from down there overshoots in the other direction. No input-reduction step in this procedure is taken below range.)
YES → block 12 · NO → block 13

Block 12 — BELOW TARGET RANGE AND RISING: the level is recovering toward range, and no input reduction is safe from here. Leave the dosing program unchanged, keep testing on the comparable schedule, and log the trend in your journal. If the rise continues PAST the top of range, this procedure reclassifies it at the range check. (Note: The verification failure path re-enters at the range check, so a rise that overshoots the range is dispositioned fresh, not waved through.) (Verify — 3 days, ALKALINITY: Alkalinity inside the target range, or still below range and approaching it, across comparable tests)
verified resolved → done · problem continues → block 11

Block 13 — Does the tank have ANY active alkalinity input: dosing pump, scheduled manual dosing, or kalkwasser in the top-off? (Note: A confirmed rise with no standing input is unusual: the cause hides outside the dosing program, and this procedure checks recent system changes before escalating.)
YES → block 14 · NO → block 25

Block 14 — Is kalkwasser part of that input (saturated limewater in the top-off, on a stirrer or doser)? (Note: Kalkwasser delivers alkalinity in proportion to evaporation, not to demand. When evaporation runs high (season change, lids off, fans on) the alkalinity input rises with it while consumption stays put.)
YES → block 15 · NO → block 16

Block 15 — KALKWASSER INPUT EXCEEDS DEMAND: evaporation is setting your alkalinity dose, and it is currently set too high. Reduce the kalkwasser concentration, or cut the share of top-off that runs through the kalk reactor, and let consumption settle the level. Do not dose anything to force it down. (Note: Reduce the input in one measured step and hold it: stacking reductions before the level responds overshoots low. There is no safe additive that lowers alkalinity.) (Verify — 3 days, ALKALINITY: Alkalinity level or falling toward the target range on the reduced kalkwasser input, across comparable tests)
verified resolved → done · problem continues → block 11

Block 16 — Is the alkalinity supply a liquid dosing pump (peristaltic or DC doser)? (Note: The delivery check below, timed container runs, quantifies a liquid dosing pump only. It cannot verify manual dosing consistency.)
YES → block 18 · NO → block 17

Block 17 — UNSUPPORTED SUPPLY METHOD — this revision of ALK-002 can verify delivered alkalinity for liquid dosing pumps only. With manual dosing, a delivery overshoot (heavy measuring, doubled doses) cannot be excluded or confirmed here, so this procedure stops rather than guess. Log your supply method and the confirmed rise in your journal; a future revision will add checks for manual dosing. (Note: If several parameters are drifting at once, not just alkalinity, a full ICP analysis may be warranted on its own merits. That is a separate decision, not this procedure's exit.) (Why this procedure stopped: ALK-002 has no verification procedure for manual dosing: the timed-container check quantifies liquid dosing pumps only. This is a Manual coverage gap, not evidence of a tank fault requiring multi-parameter analysis.)
END — UNSUPPORTED METHOD

Block 18 — Verify actual delivered dose. Run your alkalinity doser into a graduated container for a timed 60-second run, twice. Convert the measured mL/day into dKH delivered with the calculator, then return here.
then → block 19

Block 19 — Across BOTH timed runs, does measured pump output match the programmed dose within ±10%? (Note: ±10% is BioForge equipment-check policy. A single run must not isolate a delivery fault: output varies run to run, and an over-delivering doser is this fault's most direct cause.)
YES → block 21 · NO → block 20

Block 20 — DOSING DELIVERY MISMATCH: the tank is not receiving the programmed dose, and over-delivery alone can produce this rise. Recalibrate the doser or replace fatigued tubing, then reprogram to the verified output and re-run the timed-container check. (Note: Verification here is the doser's measured output: the alkalinity trend is re-checked downstream once delivery is proven. Recalibrate on a schedule, not on symptoms.) (Verify — 1 day, DOSER OUTPUT: Two fresh timed 60-second container runs each match the programmed dose within ±10%)
verified resolved → done · problem continues → block 11

Block 21 — Has calcification demand visibly dropped since the dose was set (coral stress or recession, coralline growth stalling, or livestock removed) while the dosing program stayed unchanged? (Note: Alkalinity demand follows calcification. When growth slows and the dose does not, the surplus accumulates as a steady rise. Treat a visible demand drop as corroboration of a dose-exceeds-demand fault, never proof.)
YES → block 22 · NO → block 25

Block 22 — Is alkalinity currently ABOVE the target range? (Note: Above range means two tasks: reduce the dose AND let the surplus settle. In range means one: trim the dose to demand before it leaves range.)
YES → block 24 · NO → block 23

Block 23 — DOSE EXCEEDS DEMAND: the tank consumes less alkalinity than the program delivers, and the level is still in range. Trim the MAINTENANCE dose to measured demand: enter tank volume and the measured rise between comparable tests in the dosing calculator, and reduce the programmed dose by the surplus. (Note: This corroborates falling demand after test error, confounds, kalkwasser, and delivery were eliminated. Trim in one measured step and re-test: stacked cuts overshoot low.) (Verify — 3 days, ALKALINITY: Comparable tests show alkalinity holding inside the target range on the trimmed maintenance dose)
verified resolved → done · problem continues → block 25

Block 24 — DOSE EXCEEDS DEMAND AND THE LEVEL IS ABOVE RANGE: two separate tasks. First trim the maintenance dose to measured demand (tank volume + measured rise between comparable tests, in the dosing calculator). Then let consumption and routine water changes settle the surplus down into range. Do not dose anything to force it down. (Note: The trimmed dose stops the rise; consumption clears the surplus. Neither substitutes for the other, and there is no safe additive that lowers alkalinity.) (Verify — 3 days, ALKALINITY: Alkalinity falling toward or back inside the target range on the trimmed dose, with no upward correction in the interval)
verified resolved → done · problem continues → block 25

Block 25 — In the last two weeks: new rock, fresh aragonite sand, cured or curing cement work, or any media change in the system? (Note: Fresh cementitious rockwork and some new media leach alkalinity as they cure: a standing input that appears on no dosing schedule.)
YES → block 26 · NO → block 27

Block 26 — RECENT SYSTEM CHANGE UNRESOLVED — a change in the last two weeks coincides with the rise, but this procedure cannot attribute the rise to it or quantify it from alkalinity readings. A full 38-parameter ICP analysis with photo and history context is the next step; note the recent change when you submit. (Why this procedure stopped: A recent system change is an unresolved confounder: curing rockwork or new media can leach alkalinity on no schedule this procedure can verify, and a confirmed rise should not be waited out on an assumption.)
END — ESCALATE DECODE

Block 27 — This procedure cannot isolate the fault from alkalinity readings alone: the standing inputs do not account for the confirmed rise. A full 38-parameter ICP analysis with photo and history context is the next step. (Why this procedure stopped: Confound window, input, delivery, and demand checks completed without isolating a cause: a confirmed alkalinity rise that the tank's inputs do not account for needs multi-parameter analysis, not another guess.)
END — ESCALATE DECODE