A pair of simultaneous exponential equations can conceal an ordinary quadratic in two positive numbers. If 2^x+2^y and 2^(x+y) are known, set p=2^x and q=2^y. Then p+q is the supplied sum and pq is the supplied exponential product. The two positive numbers are roots of a quadratic; logarithms recover their exponents.
For example, solve 2^x+2^y=10 and 2^(x+y)=16. Put p=2^x>0, q=2^y>0. We have p+q=10 and pq=16, so p and q are roots of z²−10z+16=0. The roots are z=2 and z=8, giving (x,y)=(1,3) or (3,1).
Did you know that some seemingly reasonable pairs of exponential sum and product cannot exist? For positive p and q, the arithmetic–geometric mean inequality requires their sum to be at least twice the square root of their product. This guide explains the method, why the answer comes in ordered pairs, how to handle negative exponents and how to test whether a parameter system has zero, one or two real ordered solutions.
Recognise a sum and a product in the original equations
The equation 2^x+2^y=10 gives a sum of two related positive quantities. The equation 2^(x+y)=16 gives their product because the index law 2^(x+y)=2^x·2^y is valid.
These two facts strongly resemble the sum and product of roots of a quadratic. There is no need to solve directly for x and y by guessing integer powers or taking logarithms of a sum.
The essential structural step is to regard 2^x and 2^y as the quantities being recovered first. Once their values are known, each exponent follows from an ordinary exponential or logarithmic equation.
This approach reuses familiar quadratic algebra while preserving the fact that exponentials of a positive base are strictly positive.
The positivity restriction comes from the exponential base
For every real x, 2^x>0. This includes negative x, fractional x and zero, but the output can never be zero or negative.
Therefore if we set p=2^x and q=2^y, then both p and q must be positive. A quadratic root that is negative is not an admissible exponential value.
Likewise, their product pq must be positive, and their sum p+q must be positive. If a proposed system demands a nonpositive sum or product, it has no real solution under this model.
State these constraints when introducing p and q. They are essential when parameter values lead to algebraic roots that exist over the reals but do not correspond to real exponents.
Translate the first equation into p plus q
Define p=2^x and q=2^y. The equation 2^x+2^y=10 becomes p+q=10 immediately.
This is the sum of the two positive substituted values. It is not the same as 2^(x+y)=10; addition and multiplication of exponentials obey different laws.
For example, 2^1+2^3=2+8=10, whereas 2^(1+3)=2^4=16.
Keeping these operations separate prevents one of the most common index-law errors. The sum equation supplies one kind of information, and the second equation will supply the product.
Translate the second equation into pq
The second condition is 2^(x+y)=16. Apply the valid exponent law 2^(x+y)=2^x·2^y.
Under p=2^x and q=2^y, this becomes pq=16. Thus the two positive numbers have sum 10 and product 16.
The same information could be written x+y=4 by taking base-two logarithms of 16, but that does not yet give x and y individually.
The p,q representation is more useful because both known relationships are now standard elementary symmetric expressions. We can find the two substituted numbers through a quadratic without further exponential manipulation.
Build the quadratic whose roots are p and q
If two numbers have sum S and product P, the monic polynomial with those numbers as roots is z²−Sz+P.
This follows by expanding (z−p)(z−q)=z²−(p+q)z+pq. The two coefficients are exactly the given sum and product.
For p+q=10 and pq=16, build z²−10z+16=0.
The variable z is temporary. It represents either p or q, not x or y. Solving this quadratic will identify the two exponential outputs without assigning which belongs to x until the final step.
Solve the hidden quadratic
Factorise z²−10z+16=(z−2)(z−8). Therefore z=2 or z=8.
Both roots are strictly positive, so both can represent 2 raised to a real power. Their sum is 2+8=10 and their product is 2·8=16.
The pair {p,q} is thus {2,8}. These are an unordered pair of positive exponential values.
At this stage, do not simply state x=2 and y=8. The numbers 2 and 8 are values of p and q, which were defined as exponentials. The original unknowns are the exponents, and the correct return step remains.
Return from p and q to x and y
If p=2 and q=8, then 2^x=2 and 2^y=8. Hence x=1 and y=3.
If p=8 and q=2, then x=3 and y=1. Both assignments satisfy the same symmetric sum and product equations.
Therefore the ordered solution pairs are (x,y)=(1,3) and (3,1).
Because the original equations treat x and y symmetrically, exchanging their roles leaves both conditions unchanged. Unless the problem adds x For x=1 and y=3, compute 2^1+2^3=2+8=10. The second equation gives 2^(1+3)=2^4=16. Both original equations are satisfied exactly. The individual exponents are real, and all exponential values are positive as required. This direct check verifies the complete ordered solution rather than merely checking the auxiliary quadratic. It also helps detect a frequent error in which the numerical values p and q are mistakenly reported as the exponents themselves. For x=3 and y=1, the first equation becomes 2^3+2^1=8+2=10. The second becomes 2^(3+1)=2^4=16. Thus the swapped solution is also valid. Both results are distinct ordered pairs because the first coordinate of one differs from the first coordinate of the other. If a word problem assigns x and y to different people, quantities or times, the order may have a practical meaning. A symmetric mathematical system alone cannot distinguish the labels, but an additional real-world condition might. Suppose the same system is accompanied by x Only (1,3) satisfies 1<3. Therefore the final answer under the added condition is x=1, y=3. Similarly, x>y would select (3,1), while x=y would make this particular system impossible because the two substituted roots are different. The correct method is to recover all pairs from the symmetric equations and then apply the extra condition. Never discard the swapped solution before reading whether the problem explicitly orders the unknowns. Solve 3^x+3^y=12 and 3^(x+y)=27. Let p=3^x>0 and q=3^y>0. Then p+q=12 and pq=27. The root quadratic is z²−12z+27=0, which factorises as (z−3)(z−9)=0. Thus {p,q}={3,9}. Returning to base-three powers gives {x,y}={1,2}. The ordered solutions are (1,2) and (2,1). Check the product: 3^(1+2)=3³=27. The method does not depend on using base two; any fixed positive base different from one works with appropriate logarithmic conversion. Let a>0, a≠1. Suppose a^x+a^y=S and a^(x+y)=P, where S and P are known real numbers. Set p=a^x>0 and q=a^y>0. Then p+q=S and pq=P. The two positive outputs, if they exist, must be roots of z²−Sz+P=0, giving z=[S±√(S²−4P)]/2. Each positive root z corresponds to exponent log_a z. Therefore the ordered pairs are formed by assigning the two positive roots in both possible orders, unless they coincide. This formula is a framework, not permission to ignore reality and positivity restrictions. The quadratic z²−Sz+P has discriminant Δ=S²−4P. If Δ<0, the quadratic has no real roots, so the exponential system has no real solutions. If Δ=0, the two substituted outputs coincide. If Δ>0, it has two distinct real roots, but we must still check that they are both strictly positive. The discriminant is therefore necessary but not by itself sufficient for a real exponential solution. A system with negative sum may have a positive discriminant and negative roots, which exponentials cannot attain. Assume S>0 and P>0, and suppose the discriminant is nonnegative. The two real roots have positive product, so they have the same sign. Their sum is S>0, so both must be positive. Therefore under S>0 and P>0, the additional condition S²≥4P guarantees admissible positive roots of the transformed quadratic. These conditions are both necessary and sufficient for the existence of real exponents x,y with the specified positive-base system. This result converts a potentially complicated exponential existence problem into a precise algebraic inequality. For p,q>0, the arithmetic–geometric mean inequality states (p+q)/2≥√(pq). Substitute p+q=S and pq=P to obtain S≥2√P. This requires S>0 and P>0. Squaring the nonnegative sides yields S²≥4P, exactly the discriminant condition from the root quadratic. The two approaches reinforce one another. The quadratic discriminant expresses the existence of two real values; AM–GM expresses the minimum sum of two positive values with fixed product. Both encode the same underlying geometry of the sum–product system. Consider 2^x+2^y=8 and 2^(x+y)=16. Then p+q=8 and pq=16. The quadratic is z²−8z+16=(z−4)²=0. Both substituted values must equal four. Thus 2^x=2^y=4, giving x=y=2. The sole distinct ordered pair is (2,2). There is no second different pair obtained by swapping, because (2,2) is unchanged. This boundary case corresponds to equality in AM–GM, where p=q=√P and the sum reaches its minimum for the fixed product. Suppose 2^x+2^y=6 while 2^(x+y)=16. Then p+q=6 and pq=16. The root quadratic is z²−6z+16, with discriminant 36−64=−28. There are no real p,q satisfying both conditions. AM–GM gives the same conclusion: two positive numbers with product 16 must have sum at least 2√16=8, yet the system demands sum six. The correct answer is no real solution. Do not search for complex exponents or approximate pairs when the problem explicitly asks for real x and y. The necessary positive-number inequality settles the question. For any real exponents and valid positive base a, the value a^(x+y) is strictly positive. Therefore a system demanding a^(x+y)=0 or a^(x+y)<0 has no real solution. Even if a quadratic with zero or negative product could be formed algebraically, that would not produce two strictly positive substituted exponential values. This is a domain check that can be done before any discriminant calculation. Identifying impossible signs early saves time and shows that the student understands the range of an exponential function rather than treating it as an unrestricted algebraic symbol. Consider 2^x+2^y=3/4 and 2^(x+y)=1/8. Set p=2^x>0 and q=2^y>0. The quadratic is z²−(3/4)z+1/8=0. Multiply by eight: 8z²−6z+1=0, or (4z−1)(2z−1)=0. Thus z=1/4 or z=1/2. Returning to exponents gives x=−2 and y=−1, or the swapped pair. Both negative exponents are permitted. They produce positive fractional exponential outputs, which satisfy the original sum and product exactly. Negative x or y is not a logarithm-domain violation. Let a=1/2. For real x, a^x is strictly positive and one-to-one, though it decreases as x increases. The same substituted algebra works. A positive transformed value p corresponds uniquely to x=log_(1/2)p. The order of exponent values may reverse compared with base two, but the existence criterion in S and P is unchanged because p,q remain positive. For example, with p=2 and q=8, the corresponding exponents are x=−1 and y=−3, or the swapped pair. Their exponential sum is still ten and their product sixteen. The base affects returned exponents, not the sum–product quadratic. Consider 2^x+2^y=7 and 2^(x+y)=10. Then p+q=7 and pq=10. The quadratic z²−7z+10 factors as (z−2)(z−5)=0. Thus the positive outputs are two and five. One exponent is log₂2=1, while the other is log₂5, which need not be an integer. The ordered solutions are (1,log₂5) and (log₂5,1). Exact logarithmic notation is an appropriate final answer. Do not assume that exponential simultaneous equations must have integer exponents simply because many school examples use small whole-number powers. The equations p+q=S and pq=P remain unchanged if p and q are exchanged. Therefore for two distinct positive quadratic roots, the data determine the unordered set of values but not their original labels. The same ambiguity transfers to x and y because a^x and a^y are one-to-one transformations of the exponents. Two distinct ordered solutions result unless an additional condition distinguishes the variables. This is not a failure of the quadratic method. It is an information-sufficiency property of the original symmetric system. A student should explain the two permutations rather than arbitrarily selecting one as the sole solution. Suppose a problem gives 2^x+2^y=10 and x+y=4, instead of 2^(x+y)=16. The second condition immediately implies 2^(x+y)=2⁴=16, so p q=16. The same sum–product quadratic z²−10z+16=0 follows. Therefore the ordered exponent solutions are again (1,3) and (3,1). This variation shows that a linear condition on exponents can supply product information for their exponential outputs. Students should look for an index law that connects the forms before using a more complicated simultaneous-equation technique. The equation 2^x+2^y=10 does not imply x+y=log₂10. The valid product law is log₂(AB)=log₂A+log₂B for positive A and B. There is no corresponding rule log₂(A+B)=log₂A+log₂B. To see the failure, take the known solution x=1,y=3. The sum x+y is four, whereas log₂10 is approximately 3.322. Applying the false logarithm-of-a-sum rule would contradict a verified exact solution. Use the substituted positive outputs instead of attempting to separate logarithms across addition. The number pair p,q={2,8} has sum ten and product sixteen. The exponent pair x,y={1,3} has sum four and product three. These are different symmetric relationships. A polynomial z²−10z+16 has roots p and q, not x and y. To obtain x and y, we apply the inverse exponential mapping after solving that polynomial. Constructing a quadratic with x+y=10 and xy=16 would be based on information the original equations never supplied. This distinction is a fundamental part of substitution. The symbols representing transformed quantities must not be silently identified with the original unknowns. Fix a product P>0 and write q=P/p for p>0. Then their sum is S(p)=p+P/p. The graph of S(p) has its unique minimum 2√P at p=√P. For a target S above that minimum, a horizontal line meets the graph twice. The two p-values are reciprocal relative to P: if one is p₁, the other is P/p₁. They correspond to exchanging p and q. At the minimum, both values coincide. Below it, there is no intersection. This provides a continuous geometric interpretation of the zero-, one- and two-solution classification from the quadratic discriminant. Suppose x+y=c, with real fixed c, and a^x+a^y=S for a>0, a≠1. The exponential product is P=a^c>0. Therefore AM–GM requires S≥2√(a^c)=2a^(c/2). If S is smaller than that threshold, there is no real ordered pair. At equality, x=y=c/2 gives one distinct pair. For S above the threshold, two distinct ordered pairs exist, obtained by solving z²−Sz+a^c=0 and taking base-a logarithms of the positive roots. This exact condition is useful for parameter questions without needing explicit exponents. The basic method assumes x and y are real. If the question additionally requires integers, the two positive substituted roots must be integer powers of the chosen base. For example, 2^x+2^y=7 and 2^(x+y)=10 has valid real solutions because the transformed outputs are two and five. But log₂5 is not an integer. Thus there are no integer exponent pairs satisfying that system, even though two real ordered solutions exist. Always separate algebraic reality from an imposed discrete condition. A problem involving days, counts or integer indices may require one final admissibility check after the logarithms are calculated. Once candidate exponents have been found, compute a^x and a^y. Check their numerical sum against S. Then check their product against P or equivalently compute a^(x+y). These are two independent requirements. For the opening pair (1,3), the sum is 2+8=10 and the product is 2·8=16. For the fractional example (−2,−1), the sum is 1/4+1/2=3/4 and the product is 1/8. Both equations must hold. Checking just the sum is insufficient, because many positive pairs share the same sum but have different products. First set p=a^x and q=a^y, explicitly stating p,q>0. Use the index law to convert the given equations into a sum S and product P. Second form the quadratic z²−Sz+P=0. Solve it exactly, checking discriminant and positivity. Third assign the positive roots to p and q in every allowed order. Fourth recover x=log_a p and y=log_a q, and apply any stated ordering, interval or integer restrictions. Finally verify both original exponential equations. The process is short because it uses structural information rather than guessing powers or incorrectly taking logarithms of a sum. Question one: 2^x+2^y=10, 2^(x+y)=16. Answers: (1,3) and (3,1). Question two: 3^x+3^y=12, 3^(x+y)=27. Answers: (1,2) and (2,1). Question three: 2^x+2^y=8, 2^(x+y)=16. Answer: (2,2) only. Question four: 2^x+2^y=6, 2^(x+y)=16. Answer: no real solution. Question five: 2^x+2^y=3/4, 2^(x+y)=1/8. Answers: (−1,−2) and (−2,−1). Question six: 2^x+2^y=7, 2^(x+y)=10. Answers: (1,log₂5) and (log₂5,1). Question seven: for x+y=c and a^x+a^y=S, give the real-solvability condition. Answer: S≥2a^(c/2), for valid base a. For each, identify which quadratic roots represent exponentials and which values are the actual exponents. The exponential product equation can provide the sum x+y directly. For the opening example, 2^(x+y)=16 gives x+y=4. The two solutions (1,3) and (3,1) have difference −2 and +2 respectively, so |x−y|=2. More generally, after obtaining p and q, calculate x−y=log_a(p/q) in the chosen order. The magnitude of that difference measures how far apart the exponents are while their sum is fixed. This can be useful if a follow-up asks for the larger exponent, their absolute difference, or an ordered pair under x For positive p and q with product P, equality in p+q≥2√P occurs exactly when p=q=√P. If p=a^x and q=a^y for a valid base a, the exponential function is one-to-one, so p=q implies x=y. Thus the repeated quadratic root case is not just an algebraic coincidence. It represents the symmetric configuration where the two exponents are equal. For 2^x+2^y=8 and 2^(x+y)=16, equality holds, giving p=q=4 and x=y=2. There is exactly one ordered pair because swapping equal values does not produce a new solution. Suppose the opening system also states x+y=4. This adds no new information because 2^(x+y)=16 already implies x+y=4 for real exponents and base two. Both candidate ordered pairs satisfy the third condition, so it does not distinguish between them. By contrast, an extra condition x−y=2 selects (3,1), while x−y=−2 selects (1,3). This illustrates independence of data. A third equation may be redundant or genuinely informative depending on whether it follows from the existing equations. Counting equations is not the same as counting independent constraints. Simultaneous exponential equations link index laws, substitution, quadratic root identities, logarithmic inverses and AM–GM inequalities. Some general parameter classifications can be enrichment depending on a learner’s school programme. Singapore families can consult the applicable SEAB syllabus resources for 2026 O-Level Additional Mathematics or 2027 SEC G3, as relevant. Related eduKateSG guides include How to Solve an Exponential Equation with Its Reciprocal, How to Find Quadratic Root Sums Without Solving for the Roots and How to Find When Two Exponential Models Become Equal. Explore the Additional Mathematics archive or enquire about suitable eduKate Singapore three-student tutorials near Sixth Avenue MRT. The central skill is recognising a positive sum–product system hidden inside exponential notation and returning every valid transformed solution to the actual exponents. Properly taught kids shine a bright light into the future.Verify the first ordered solution
Verify the swapped ordered solution
An extra ordering condition selects one answer
A second example with base three
The general solution uses positive sum and product
The discriminant decides whether real p and q exist
Positive sum and product give positive roots when real
The arithmetic–geometric mean gives the same condition
A repeated root produces one ordered pair
An impossible system despite positive right-hand sides
A nonpositive target product is impossible
Negative exponents are valid and can produce fractional sums
A base below one still gives a valid one-to-one mapping
Recover exponents when the quadratic roots are not neat powers
The sum and product do not tell us which label belongs to which root
A sum-of-exponents condition provides the same product information
Do not take the logarithm of a sum term by term
Why the quadratic is built from the output values, not the exponent values
A graphical view of the two-solution symmetry
Parameter classification with a fixed exponent sum
An integer-only restriction changes the problem
A quick verification method based on the index laws
A concise exam solution routine
Practice problems with checked answers
Recover the sum and difference of the exponents
Why equality in AM–GM forces equal exponents
Use a third equation only as a consistency check
Syllabus alignment and connected reading